Resident Evil
← all movies. Each comment's label under every pathway (+ positive, − negative, 0 neutral / mixed / no verdict; hover a cell for the verdict).
Pick a pathway to show its verdict and rationale under every comment: VADER (old v1) · siebert+opinion (previous live) · Local NLI stance (deberta-v3-large) · GPT-6.1 Sol (LIVE) · Claude Opus 5.5 · Claude Sonnet 5.5 · Gemini 4 Argon · Grok 4.7 · GPT-6 Luna (cheap) · Claude Haiku 5.5 (cheap) · Gemini 3.5 Flash-Lite (cheap)
| comment (excerpt ≤280 chars) | VADER (old v1) | siebert+opinion (previous live) | Local NLI stance (deberta-v3-large) | GPT-6.1 Sol (LIVE) | Claude Opus 5.5 | Claude Sonnet 5.5 | Gemini 4 Argon | Grok 4.7 | GPT-6 Luna (cheap) | Claude Haiku 5.5 (cheap) | Gemini 3.5 Flash-Lite (cheap) |
|---|---|---|---|---|---|---|---|---|---|---|---|
| rank · score · bucket · Reddit · comments +/0/− | #9 71 Depends R 55 48/12/36 | #7 72 Depends R 58 36/32/28 | #2 73 Depends R 59 23/55/18 | #4 91 Watch R 98 42/53/1 | #3 92 Watch R 100 37/59/0 | #5 89 Depends R 94 36/58/2 | #3 92 Watch R 100 41/55/0 | #5 90 Depends R 96 38/56/2 | #3 91 Watch R 97 45/49/2 | #1 92 Watch R 100 34/62/0 | #2 92 Watch R 100 40/56/0 |
| thread #1 w 0.707 Loved it. I really like the idea that you are following someone who is ultimately just a victim like anyone else in the city, complete with an audio log left behind. It was a little abrupt at the end but that's ok cause I was fully invested the entire runtime. I also love the… VADER (old v1): positive — VADER 0.901 siebert+opinion (previous live): positive — P(op)=0.95 P(pos)=1.00 Local NLI stance (deberta-v3-large): positive — NLI p=0.91 GPT-6.1 Sol (LIVE): positive — Explicitly loved it, stayed fully invested, and enjoyed the humor despite finding the ending slightly abrupt. Claude Opus 5.5: positive — Says 'loved it', was fully invested; minor complaint about abrupt ending doesn't outweigh praise. Claude Sonnet 5.5: positive — Says 'Loved it', was fully invested throughout; the only criticism is a minor note about an abrupt ending, which they dismiss. Gemini 4 Argon: positive — The commenter explicitly states they 'Loved it' and were 'fully invested the entire runtime' despite an abrupt ending. Grok 4.7: positive — Loved it and stayed fully invested despite an abrupt ending. GPT-6 Luna (cheap): positive — They loved it and were fully invested throughout; the abrupt ending is a minor caveat. Claude Haiku 5.5 (cheap): positive — Opens with 'Loved it', praises the premise and runtime investment; the abrupt ending is a minor caveat. Gemini 3.5 Flash-Lite (cheap): positive — The commenter states 'Loved it' and describes being fully invested despite minor flaws. | + | + | + | + | + | + | + | + | + | + | + |
| thread #2 w 0.577 I know ppl hated the ending but I thought it was cool we basically got the orgin story of a boss Leon has to fight VADER (old v1): negative — VADER -0.115 siebert+opinion (previous live): positive — P(op)=0.84 P(pos)=0.99 Local NLI stance (deberta-v3-large): mixed — NLI p=0.73 GPT-6.1 Sol (LIVE): positive — They defend the ending as cool, appreciating its origin story for a boss Leon fights. Claude Opus 5.5: positive — Commenter defends the disliked ending, calling it cool as a villain origin story. Claude Sonnet 5.5: positive — Commenter defends the disliked ending, calling it cool and enjoying the origin story of a boss, indicating a favorable view. Gemini 4 Argon: positive — The commenter explicitly praises the film's ending, calling the boss origin story cool despite others hating it. Grok 4.7: positive — Liked the ending as a cool origin story despite others hating it. GPT-6 Luna (cheap): positive — They thought the ending was cool, expressing a positive view of the film. Claude Haiku 5.5 (cheap): mixed — Acknowledges others hated the ending but personally thought it was cool; a mild positive lean on one element, overall unclear. Gemini 3.5 Flash-Lite (cheap): positive — The commenter liked the ending and thought the origin story element was cool. | − | + | 0 | + | + | + | + | + | + | 0 | + |
| thread #3 w 0.5 Holy crap yes! He IS that audio log guy! Lol VADER (old v1): positive — VADER 0.54 siebert+opinion (previous live): neutral — P(op)=0.18 P(pos)=1.00 Local NLI stance (deberta-v3-large): no_verdict — NLI p=0.47 GPT-6.1 Sol (LIVE): no_verdict — Recognizes someone as the audio log guy without judging the film. Claude Opus 5.5: no_verdict — Recognizes an actor or character connection; no judgement of the film's quality. Claude Sonnet 5.5: no_verdict — Commenter is excitedly recognizing an actor as the audio log guy; this is trivia with no judgement of the film. Gemini 4 Argon: no_verdict — The comment recognizes an actor or character reference without expressing any opinion on the movie itself. Grok 4.7: no_verdict — Recognizes a character from an audio log; no film judgement. GPT-6 Luna (cheap): no_verdict — Recognizes an actor or character from an audio log; expresses no opinion about the film. Claude Haiku 5.5 (cheap): no_verdict — Comment notes a trivia fact about an actor appearing in an audio log, with no judgement of the film's quality. Gemini 3.5 Flash-Lite (cheap): no_verdict — The comment is a joke/reaction about an actor, not a verdict on the film. | + | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| thread #4 w 0.447 Scariest moment of the whole movie was Brian waving the gun barrel in front of Paul Walter Hauser’s face in the car while loading it VADER (old v1): negative — VADER -0.34 siebert+opinion (previous live): negative — P(op)=0.98 P(pos)=0.00 Local NLI stance (deberta-v3-large): positive — NLI p=0.60 GPT-6.1 Sol (LIVE): no_verdict — Identifies the scariest moment without expressing an overall opinion of the film. Claude Opus 5.5: no_verdict — Joke about unsafe gun handling in a scene; no clear judgement of the film's quality. Claude Sonnet 5.5: positive — Praises a moment as the scariest of the movie; scariness is treated as a compliment, implying a favorable view of the film. Gemini 4 Argon: no_verdict — The comment makes a joke about a specific scene without expressing an overall opinion on the film. Grok 4.7: no_verdict — Notes one scary scene without judging the film overall. GPT-6 Luna (cheap): positive — Calling it the movie’s scariest moment praises its effectiveness as a scare. Claude Haiku 5.5 (cheap): no_verdict — Describes a specific scene as the scariest moment without clearly judging the film's overall quality or recommending it. Gemini 3.5 Flash-Lite (cheap): no_verdict — Discusses a specific scary moment without expressing an overall opinion on the film. | − | − | + | 0 | 0 | + | 0 | 0 | + | 0 | 0 |
| thread #5 w 0.408 "YOU COULD SHOOT THE ENGINE!" VADER (old v1): negative — VADER -0.4 siebert+opinion (previous live): negative — P(op)=0.87 P(pos)=0.00 Local NLI stance (deberta-v3-large): negative — NLI p=0.30 GPT-6.1 Sol (LIVE): no_verdict — The comment quotes dialogue without expressing an opinion about the film. Claude Opus 5.5: no_verdict — Quotes a line from the film without expressing any judgement of the movie. Claude Sonnet 5.5: no_verdict — A quoted line from the film (likely mocking or referencing a scene) with no explicit judgement of the film itself. Gemini 4 Argon: no_verdict — The comment is simply a quote or reaction to a specific scene without expressing an overall opinion on the film. Grok 4.7: no_verdict — Quoted line only; no judgment of the film. GPT-6 Luna (cheap): no_verdict — A quoted line with no judgment of the film. Claude Haiku 5.5 (cheap): no_verdict — Just quotes a line from the film without any judgement of the movie itself. Gemini 3.5 Flash-Lite (cheap): no_verdict — Just a movie quote, no judgment of the film. | − | − | − | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| thread #6 w 0.378 Legit thought that's how he was gonna die. By getting shot accidentally. VADER (old v1): negative — VADER -0.743 siebert+opinion (previous live): neutral — P(op)=0.14 P(pos)=0.00 Local NLI stance (deberta-v3-large): no_verdict — NLI p=0.41 GPT-6.1 Sol (LIVE): no_verdict — Discusses an expected character death without judging the film. Claude Opus 5.5: no_verdict — Comment speculates about a character's death without judging the film's quality. Claude Sonnet 5.5: no_verdict — Comment is about a plot expectation regarding a character's death, with no judgment of the film's quality. Gemini 4 Argon: no_verdict — The comment only discusses a specific plot point and character's fate without expressing an opinion on the film's quality. Grok 4.7: no_verdict — Scene speculation only; no judgment of the film. GPT-6 Luna (cheap): no_verdict — They speculate about how a character might die, without judging the film. Claude Haiku 5.5 (cheap): no_verdict — Speculation about a character's death; no judgement of the film's quality or recommendation. Gemini 3.5 Flash-Lite (cheap): no_verdict — Comment discusses a plot expectation without judging the film itself. | − | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| thread #7 w 0.354 Can’t get over the dynamic lighting in the first half of the movie. The way the snow shined in blood red tail lights with a suffocating pitch black background was so cool. Top notch cinematography and man Cregger knows how to build the suspense while balancing such a great… VADER (old v1): positive — VADER 0.955 siebert+opinion (previous live): positive — P(op)=0.98 P(pos)=1.00 Local NLI stance (deberta-v3-large): positive — NLI p=0.95 GPT-6.1 Sol (LIVE): positive — Explicitly loved the film, praising its cinematography, suspense, lighting, and comedy. Claude Opus 5.5: positive — Praises lighting, cinematography, suspense and comedy balance; explicitly says they absolutely loved it. Claude Sonnet 5.5: positive — Praises the cinematography, suspense, and comedic balance, and says explicitly 'Absolutely loved it.' Gemini 4 Argon: positive — The commenter explicitly states they 'Absolutely loved it' and praises the cinematography, lighting, suspense, and comedy. Grok 4.7: positive — Praises cinematography, suspense, comedy, and says they absolutely loved it. GPT-6 Luna (cheap): positive — They praise the cinematography, suspense, and comedy, and explicitly say they absolutely loved the film. Claude Haiku 5.5 (cheap): positive — Praises the cinematography, suspense, and comedy, and says "Absolutely loved it." Gemini 3.5 Flash-Lite (cheap): positive — The commenter explicitly states 'Top notch cinematography' and 'Absolutely loved it'. | + | + | + | + | + | + | + | + | + | + | + |
| thread #8 w 0.333 I thought the same thing in the beginning with the shadows on the white snow working so perfect. The red from the break lights and cop car were also perfect with adding the tension fear and anxiety. It was really well done. VADER (old v1): positive — VADER 0.684 siebert+opinion (previous live): positive — P(op)=0.95 P(pos)=1.00 Local NLI stance (deberta-v3-large): positive — NLI p=0.75 GPT-6.1 Sol (LIVE): positive — Praises the lighting and visual choices for effectively creating tension, fear, and anxiety. Claude Opus 5.5: positive — Praises the visuals, lighting and tension-building as perfect and really well done. Claude Sonnet 5.5: positive — Praises the cinematography (shadows on snow, red lights) as perfect and says the film was really well done. Gemini 4 Argon: positive — The commenter praises the visual choices, tension, and atmosphere, calling it 'perfect' and 'really well done'. Grok 4.7: positive — Praises the scene's visuals and tension as really well done. GPT-6 Luna (cheap): positive — Praises the visual lighting and color choices, saying they effectively build tension and are really well done. Claude Haiku 5.5 (cheap): positive — Praises the film's visual craft, calling the shadows, red lights and tension 'perfect' and 'really well done'. Gemini 3.5 Flash-Lite (cheap): positive — The commenter praises the visual style and tension, stating it was really well done. | + | + | + | + | + | + | + | + | + | + | + |
| thread #9 w 0.316 Shooting it like a video game was a great choice. Felt like we were in the middle of the action and story VADER (old v1): positive — VADER 0.844 siebert+opinion (previous live): positive — P(op)=0.44 P(pos)=1.00 Local NLI stance (deberta-v3-large): no_verdict — NLI p=0.45 GPT-6.1 Sol (LIVE): positive — Praises the video-game shooting style for immersing viewers in the action and story. Claude Opus 5.5: positive — Praises the video-game filming style as a great choice that felt immersive. Claude Sonnet 5.5: positive — Praises the video-game-style cinematography as a great choice that made the action and story immersive. Gemini 4 Argon: positive — The commenter praises the stylistic choice and found the action and story immersive. Grok 4.7: positive — Praises the video-game style shooting and immersive action. GPT-6 Luna (cheap): positive — Praises the film’s video-game-inspired style for making the action and story immersive. Claude Haiku 5.5 (cheap): positive — Commenter praises the film's video-game-style shooting as a great choice that made them feel immersed. Gemini 3.5 Flash-Lite (cheap): positive — Praises the film's style, stating it was a great choice and felt immersive. | + | + | 0 | + | + | + | + | + | + | + | + |
| thread #10 w 0.302 infected foot kid: 🙆♀️🙆♀️ Bryan: … damn .. VADER (old v1): negative — VADER -0.71 siebert+opinion (previous live): positive — P(op)=0.85 P(pos)=0.96 Local NLI stance (deberta-v3-large): negative — NLI p=0.43 GPT-6.1 Sol (LIVE): no_verdict — A joke about characters and a scene, without judging the film. Claude Opus 5.5: no_verdict — Joke referencing characters/scene with emojis; no judgement of the film's quality. Claude Sonnet 5.5: no_verdict — A joking emoji-laden reference to a scene or character with no explicit judgment of the film's quality. Gemini 4 Argon: no_verdict — The comment is a joke about a specific scene/character interaction without expressing an opinion on the film's quality. Grok 4.7: no_verdict — Scene quote/joke only; no judgment of the film. GPT-6 Luna (cheap): no_verdict — A brief joke or reaction with no clear opinion about the film. Claude Haiku 5.5 (cheap): no_verdict — Quotes a scene with a reaction but gives no clear judgement of the film itself. Gemini 3.5 Flash-Lite (cheap): no_verdict — Just quoting or referencing a specific scene with emojis, no judgment of the film. | − | + | − | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| thread #11 w 0.289 My theatre laughed in unison outloud when she said hold me and then cut to his face.. VADER (old v1): positive — VADER 0.226 siebert+opinion (previous live): negative — P(op)=0.42 P(pos)=0.01 Local NLI stance (deberta-v3-large): negative — NLI p=0.31 GPT-6.1 Sol (LIVE): no_verdict — Describes the audience laughing at a scene without expressing an overall opinion of the film. Claude Opus 5.5: no_verdict — Anecdote about the audience laughing at one scene; no overall judgement of the film is given. Claude Sonnet 5.5: negative — The theatre laughing in unison at what seems a serious moment implies unintentional comedy, mocking the film, though no explicit verdict is stated. Gemini 4 Argon: no_verdict — Describes the audience's reaction to a specific scene without stating a personal verdict on the movie overall. Grok 4.7: no_verdict — Notes audience laughter at a scene, no film judgement. GPT-6 Luna (cheap): no_verdict — Reports the audience laughing at a scene but does not state an opinion of the film overall. Claude Haiku 5.5 (cheap): no_verdict — Describes a theatre audience's reaction to a scene; no clear judgement of the film's quality or recommendation. Gemini 3.5 Flash-Lite (cheap): no_verdict — The commenter is sharing a humorous theater experience without expressing a clear verdict on the film itself. | + | − | − | 0 | 0 | − | 0 | 0 | 0 | 0 | 0 |
| thread #12 w 0.277 His resigned “Damn” really sold me on his character, he’s not a bad dude but woefully unprepared for this kind of situation. VADER (old v1): negative — VADER -0.481 siebert+opinion (previous live): negative — P(op)=0.95 P(pos)=0.00 Local NLI stance (deberta-v3-large): mixed — NLI p=0.43 GPT-6.1 Sol (LIVE): positive — Praises the character portrayal, highlighting how the resigned delivery convincingly conveyed his vulnerability. Claude Opus 5.5: positive — Praises a character moment that sold them on the character; a positive view of the film's execution. Claude Sonnet 5.5: positive — Praises an actor's delivery as selling the character, which is a favorable judgment of the film's performance, though it's about one moment rather than the whole film. Gemini 4 Argon: positive — The commenter praises a specific moment of characterization that 'sold' them on the character. Grok 4.7: no_verdict — Character moment analysis, no overall film judgement. GPT-6 Luna (cheap): positive — They praise the film’s characterization, saying the line sold them on the character. Claude Haiku 5.5 (cheap): no_verdict — Comments on a character's acting moment and his personality, with no judgement of the film overall. Gemini 3.5 Flash-Lite (cheap): no_verdict — Comment discusses a specific character and scene without judging the film overall. | − | − | 0 | + | + | + | + | 0 | + | 0 | 0 |
| thread #13 w 0.267 Her face splitting open to the single eyeball got a nice jump outta me. Was she playing him the whole time? One of my favorite scenes VADER (old v1): positive — VADER 0.765 siebert+opinion (previous live): neutral — P(op)=0.39 P(pos)=1.00 Local NLI stance (deberta-v3-large): positive — NLI p=0.60 GPT-6.1 Sol (LIVE): positive — Praises the scene as a favorite and appreciates its effective jump scare. Claude Opus 5.5: positive — Praises a scary scene as one of their favorites; the plot question doesn't negate the approval. Claude Sonnet 5.5: positive — Commenter enjoyed the jump scare and calls it one of their favorite scenes, indicating a favorable view of the film. Gemini 4 Argon: positive — The commenter praises a jump scare and calls it one of their favorite scenes. Grok 4.7: positive — Calls a jump-scare scene one of their favorites. GPT-6 Luna (cheap): positive — Calls it one of their favorite scenes and praises its jump scare. Claude Haiku 5.5 (cheap): positive — Calls the gory face-splitting scene a nice jump and names it one of their favorite scenes. Gemini 3.5 Flash-Lite (cheap): positive — The commenter praises a specific scene, calling it one of their favorites and noting it got a good jump. | + | 0 | + | + | + | + | + | + | + | + | + |
| thread #14 w 0.258 I think it’s more like how Brian was at the end; pleading with the scientists to cure him while his body is acting on the monster instincts VADER (old v1): positive — VADER 0.42 siebert+opinion (previous live): positive — P(op)=0.70 P(pos)=0.99 Local NLI stance (deberta-v3-large): no_verdict — NLI p=0.37 GPT-6.1 Sol (LIVE): no_verdict — Discusses a character’s behavior and monster instincts without judging the film. Claude Opus 5.5: no_verdict — Plot interpretation comparing a character's ending; no judgement of the film's quality. Claude Sonnet 5.5: no_verdict — Comment draws a plot/character comparison about Brian's ending, offering no judgment of the film's quality or recommendation. Gemini 4 Argon: no_verdict — The comment merely discusses a character's situation and plot details without expressing an opinion on the film's quality. Grok 4.7: no_verdict — Plot interpretation of a character's ending, no film judgement. GPT-6 Luna (cheap): no_verdict — Speculates about a character's behavior without expressing an opinion on the film. Claude Haiku 5.5 (cheap): no_verdict — Comment offers a plot interpretation of a character's ending, with no judgement of the film's quality or recommendation. Gemini 3.5 Flash-Lite (cheap): no_verdict — The comment discusses plot and character details without expressing any opinion or verdict on the film itself. | + | + | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| thread #15 w 0.25 That made it much more sad for me. We know the poor kid is screwed, but she doesn’t realize yet, and she’s just a scared little girl who wants someone to help her. VADER (old v1): negative — VADER -0.685 siebert+opinion (previous live): negative — P(op)=0.91 P(pos)=0.00 Local NLI stance (deberta-v3-large): negative — NLI p=0.39 GPT-6.1 Sol (LIVE): no_verdict — Describes sadness about a character’s situation without judging the film’s quality or recommending it. Claude Opus 5.5: no_verdict — Describes emotional impact of a character's plight without judging the film's quality. Claude Sonnet 5.5: no_verdict — Discusses emotional impact of a character's situation in a scene, but gives no explicit judgement of the film or recommendation. Gemini 4 Argon: no_verdict — Discusses the emotional context of a character's situation without expressing an opinion on the film's overall quality. Grok 4.7: no_verdict — Emotional plot reaction about a child's fate, no film judgement. GPT-6 Luna (cheap): no_verdict — They describe feeling sad for a character, but do not judge the film itself. Claude Haiku 5.5 (cheap): no_verdict — Discusses a character's emotional impact in a scene but gives no clear judgement of the film's quality or recommendation. Gemini 3.5 Flash-Lite (cheap): no_verdict — Comment discusses plot sadness and character empathy without judging the overall film quality. | − | − | − | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| thread #16 w 0.243 i dont think she was playing him , even after holding him with her foot she kept screaming for help VADER (old v1): positive — VADER 0.226 siebert+opinion (previous live): neutral — P(op)=0.05 P(pos)=0.00 Local NLI stance (deberta-v3-large): negative — NLI p=0.40 GPT-6.1 Sol (LIVE): no_verdict — Discusses a character's intentions and actions without judging the film. Claude Opus 5.5: no_verdict — Plot interpretation about a character's actions; no judgement of the film's quality. Claude Sonnet 5.5: no_verdict — Comment discusses a character's behavior in a scene, with no judgment of the film itself. Gemini 4 Argon: no_verdict — The comment is purely plot and character action discussion without any judgment of the movie itself. Grok 4.7: no_verdict — Scene interpretation only; no judgement of the film. GPT-6 Luna (cheap): no_verdict — Discusses a character's actions and motives without judging the film. Claude Haiku 5.5 (cheap): no_verdict — Discusses a plot detail about a character's actions with no judgement of the film's quality. Gemini 3.5 Flash-Lite (cheap): no_verdict — The comment discusses plot details and character actions without expressing a judgment on the film. | + | 0 | − | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| thread #17 w 0.236 Damn, that pedestrian he hit could sure hold a grudge. VADER (old v1): negative — VADER -0.103 siebert+opinion (previous live): negative — P(op)=0.73 P(pos)=0.00 Local NLI stance (deberta-v3-large): positive — NLI p=0.27 GPT-6.1 Sol (LIVE): no_verdict — A joke about a character and plot event, without judging the film. Claude Opus 5.5: no_verdict — Joke about a plot moment; no judgement of the film's quality. Claude Sonnet 5.5: no_verdict — A joke about a plot point (pedestrian hit by a character) without any judgement of the film itself. Gemini 4 Argon: no_verdict — The comment makes a joke about a plot point without expressing an opinion on the film's quality. Grok 4.7: no_verdict — Joke about a plot moment, no judgment of the film. GPT-6 Luna (cheap): no_verdict — A joke about a character's actions, with no judgment of the film itself. Claude Haiku 5.5 (cheap): no_verdict — Joking about a plot event (the pedestrian's grudge) with no judgement of the film itself. Gemini 3.5 Flash-Lite (cheap): no_verdict — A joke about a plot point, no judgment of the film. | − | − | + | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| thread #18 w 0.229 The officer: "Sweet Jesus what the hell happened to her" 😂😂 VADER (old v1): positive — VADER 0.778 siebert+opinion (previous live): neutral — P(op)=0.26 P(pos)=0.01 Local NLI stance (deberta-v3-large): positive — NLI p=0.33 GPT-6.1 Sol (LIVE): no_verdict — Quotes a character's line with laughing emojis without judging the film itself. Claude Opus 5.5: no_verdict — Quotes a line with laughing emojis; amusement at a moment, no judgement of the film overall. Claude Sonnet 5.5: no_verdict — Quotes a line from the film with laughing emojis; no explicit judgement of the film's quality or recommendation. Gemini 4 Argon: no_verdict — The comment only quotes a line from the movie with laughing emojis, expressing no overall judgment. Grok 4.7: no_verdict — Quotes a line and laughs; no judgment of the film. GPT-6 Luna (cheap): no_verdict — Quotes a character’s reaction and adds laughing emojis, without judging the film. Claude Haiku 5.5 (cheap): no_verdict — Quotes a line of dialogue with laughing emojis; no judgement of the film's quality or whether they'd recommend it. Gemini 3.5 Flash-Lite (cheap): no_verdict — Quoting a line with emojis is not a verdict on the film. | + | 0 | + | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| thread #19 w 0.224 Dude accidentally killed the officer telling him to hold her in his car VADER (old v1): negative — VADER -0.34 siebert+opinion (previous live): neutral — P(op)=0.13 P(pos)=0.00 Local NLI stance (deberta-v3-large): negative — NLI p=0.47 GPT-6.1 Sol (LIVE): no_verdict — Describes a plot event without expressing an opinion on the film. Claude Opus 5.5: no_verdict — Describes a plot moment without any judgement of the film's quality. Claude Sonnet 5.5: no_verdict — Comment only recounts a plot event and gives no judgement of the film. Gemini 4 Argon: no_verdict — The comment simply describes a plot point from the movie without expressing an opinion on the film's quality. Grok 4.7: no_verdict — Describes a plot moment without judging the film. GPT-6 Luna (cheap): no_verdict — Describes a plot event without expressing an opinion about the film. Claude Haiku 5.5 (cheap): no_verdict — Plot description of a scene with no judgement of the film's quality or recommendation. Gemini 3.5 Flash-Lite (cheap): no_verdict — The comment only discusses a plot point without expressing any opinion or verdict on the film. | − | 0 | − | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| thread #20 w 0.218 IM SORRY FOR HITTING YOU WITH MY CAR VADER (old v1): negative — VADER -0.077 siebert+opinion (previous live): neutral — P(op)=0.26 P(pos)=0.00 Local NLI stance (deberta-v3-large): mixed — NLI p=0.40 GPT-6.1 Sol (LIVE): no_verdict — The comment quotes dialogue without expressing an opinion on the film. Claude Opus 5.5: no_verdict — Quoting or joking about a line or scene; no judgement of the film's quality. Claude Sonnet 5.5: no_verdict — A joke or quote referencing a scene, with no judgement of the film's quality or recommendation. Gemini 4 Argon: no_verdict — The comment is just a quote or joke referencing a scene without expressing an opinion on the film. Grok 4.7: no_verdict — Quote or joke with no judgment of the film. GPT-6 Luna (cheap): no_verdict — A quoted apology with no opinion about the film. Claude Haiku 5.5 (cheap): no_verdict — A joke quote referencing a scene; no judgement of the film's quality or recommendation. Gemini 3.5 Flash-Lite (cheap): no_verdict — Just quoting a line from the movie, no review or verdict. | − | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| thread #22 w 0.209 Him apologizing to her like OK IM SORRY I HIT YOU WITH MY CAR was hilarious VADER (old v1): positive — VADER 0.7 siebert+opinion (previous live): neutral — P(op)=0.36 P(pos)=1.00 Local NLI stance (deberta-v3-large): no_verdict — NLI p=0.52 GPT-6.1 Sol (LIVE): positive — Calls a moment in the film hilarious, expressing enjoyment of its humor. Claude Opus 5.5: no_verdict — Quotes and laughs at one scene; no overall judgement of the film's quality. Claude Sonnet 5.5: no_verdict — Comment finds a specific scene funny, but it's a joke about plot and gives no overall verdict on the film. Gemini 4 Argon: no_verdict — The commenter finds one specific scene funny without giving an overall verdict on the film itself. Grok 4.7: positive — Finds a specific scene hilarious, a positive reaction to the film. GPT-6 Luna (cheap): positive — They call a moment in the film hilarious, expressing enjoyment of it. Claude Haiku 5.5 (cheap): no_verdict — Comment finds one scene funny but gives no overall judgement of the film's quality or recommendation. Gemini 3.5 Flash-Lite (cheap): no_verdict — The commenter shares a funny quote without expressing an overall opinion on the film. | + | 0 | 0 | + | 0 | 0 | 0 | + | + | 0 | 0 |
| thread #23 w 0.204 I appreciate how much of a dweeb the main character was. There was never a point I didn't feel he was under genuine threat because he was not built for this. VADER (old v1): negative — VADER -0.178 siebert+opinion (previous live): negative — P(op)=1.00 P(pos)=0.01 Local NLI stance (deberta-v3-large): positive — NLI p=0.34 GPT-6.1 Sol (LIVE): positive — Praises the protagonist's vulnerability for making the danger feel genuine throughout. Claude Opus 5.5: positive — Appreciates the protagonist's vulnerability, which made the threat feel genuine; praise for the film's tension. Claude Sonnet 5.5: positive — Commenter appreciates the protagonist's dweebiness, which made the threat feel genuine; this praises the film's characterization and tension. Gemini 4 Argon: positive — The commenter praises the characterization of the protagonist and how effectively it created genuine tension. Grok 4.7: positive — Appreciates the vulnerable lead making the threat feel genuine. GPT-6 Luna (cheap): positive — They appreciate the protagonist’s characterization and say it made the danger feel genuine. Claude Haiku 5.5 (cheap): positive — Commenter appreciates the character's portrayal and feels genuine threat throughout, implying approval of the film's craft. Gemini 3.5 Flash-Lite (cheap): positive — Appreciates the character's vulnerability and genuine sense of threat. | − | − | + | + | + | + | + | + | + | + | + |
| thread #24 w 0.2 I had this thought. Like, this character is not a Resident Evil protagonist. He's a player thrust into a Resident Evil situation. The "protagonist" is Maxine, who shows up, bodies zombies like they're nothing, then gets taken out by the boss the first time she tries to fight… VADER (old v1): negative — VADER -0.776 siebert+opinion (previous live): neutral — P(op)=0.33 P(pos)=0.00 Local NLI stance (deberta-v3-large): no_verdict — NLI p=0.38 GPT-6.1 Sol (LIVE): positive — Expresses appreciation for the film’s recurring typewriter joke and its clever connection to game mechanics. Claude Opus 5.5: positive — Commenter enjoys the game-meta concept and says they 'dig' the typewriter joke; mostly analytical but appreciative. Claude Sonnet 5.5: positive — Commenter enthusiastically analyzes the film's clever game-logic structure and says they 'dig' the typewriter joke, implying appreciation of the film. Gemini 4 Argon: positive — The commenter praises a specific meta-joke in the film ('I also dig the little thru-line joke') and mentions rewatching it. Grok 4.7: no_verdict — Analyzes game-like structure and a joke; no overall film judgement. GPT-6 Luna (cheap): positive — They explicitly say they enjoy the film’s recurring joke about Bryan not using the Typewriter. Claude Haiku 5.5 (cheap): no_verdict — Discusses plot structure, characters, and a game-reference joke without stating whether they liked or recommend the film. Gemini 3.5 Flash-Lite (cheap): positive — The commenter enjoys specific elements like the thru-line joke and rewatching the film. | − | 0 | 0 | + | + | + | + | 0 | + | 0 | + |
| thread #25 w 0.196 The newer games have protagonists more like him. RE 7 and RE 9 both have player characters that are just regular people basically and are terrible at things like using guns. VADER (old v1): positive — VADER 0.294 siebert+opinion (previous live): neutral — P(op)=0.06 P(pos)=0.00 Local NLI stance (deberta-v3-large): no_verdict — NLI p=0.38 GPT-6.1 Sol (LIVE): no_verdict — Discusses protagonists in the games without judging the film. Claude Opus 5.5: no_verdict — Comment compares a character to video game protagonists; no judgement of the film itself. Claude Sonnet 5.5: no_verdict — Comment discusses game protagonists and franchise trends, offering no judgement of the film itself. Gemini 4 Argon: no_verdict — The comment compares a character to the video game protagonists without expressing an opinion on the movie itself. Grok 4.7: no_verdict — Compares game protagonists; no judgment of the film. GPT-6 Luna (cheap): no_verdict — Discusses protagonists in newer games without judging the film. Claude Haiku 5.5 (cheap): no_verdict — Compares game protagonists across the franchise's newer titles; no judgement of the film itself is expressed. Gemini 3.5 Flash-Lite (cheap): no_verdict — The comment discusses video game protagonists and mechanics without judging the specific film. | + | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| thread #26 w 0.192 I also appreciated how he was just muttering profanities and ranting to himself as shit just gets worse and worse. Excellent script VADER (old v1): negative — VADER -0.421 siebert+opinion (previous live): positive — P(op)=0.85 P(pos)=1.00 Local NLI stance (deberta-v3-large): positive — NLI p=0.33 GPT-6.1 Sol (LIVE): positive — Appreciates the character's reactions and explicitly praises the script as excellent. Claude Opus 5.5: positive — Commenter appreciates a character's behavior and explicitly praises the script as excellent. Claude Sonnet 5.5: positive — Commenter appreciated a character's muttering and ranting as things escalate, and explicitly calls it an 'Excellent script'. Gemini 4 Argon: positive — The commenter praises a character's dialogue and explicitly calls the script "excellent". Grok 4.7: positive — Praises the script and a character moment as excellent. GPT-6 Luna (cheap): positive — They praise the film’s script and appreciate the character’s muttering as events worsen. Claude Haiku 5.5 (cheap): positive — Commenter praises the character's ranting and the script as 'excellent', a clear approving verdict on the film. Gemini 3.5 Flash-Lite (cheap): positive — Appreciated specific elements and explicitly praised the script as excellent. | − | + | + | + | + | + | + | + | + | + | + |
| thread #27 w 0.189 getting tired and giving up on level 4 of the stairs was soooo real VADER (old v1): negative — VADER -0.128 siebert+opinion (previous live): positive — P(op)=0.44 P(pos)=0.99 Local NLI stance (deberta-v3-large): no_verdict — NLI p=0.37 GPT-6.1 Sol (LIVE): no_verdict — Calls a character’s exhaustion relatable without judging the film overall. Claude Opus 5.5: no_verdict — A relatable joke about a scene, with no judgement of the film's quality. Claude Sonnet 5.5: no_verdict — Relatable joke about a scene or experience; no judgment of the film's quality or recommendation. Gemini 4 Argon: no_verdict — The comment makes a joke about a relatable scene without expressing an overall opinion on the film. Grok 4.7: positive — Praises a scene for feeling authentically relatable. GPT-6 Luna (cheap): no_verdict — Relates to a moment in the film but gives no opinion on the film itself. Claude Haiku 5.5 (cheap): no_verdict — Joking reference to a game-like stair level; no judgement of the film itself expressed. Gemini 3.5 Flash-Lite (cheap): no_verdict — The comment is a joke about a video game level, not a verdict on the film. | − | + | 0 | 0 | 0 | 0 | 0 | + | 0 | 0 | 0 |
| thread #28 w 0.186 That dog jump scares got me bad, and then the immediate surprise reveal 10 seconds later was amazing. VADER (old v1): neutral — VADER 0.0 siebert+opinion (previous live): positive — P(op)=0.98 P(pos)=1.00 Local NLI stance (deberta-v3-large): positive — NLI p=0.47 GPT-6.1 Sol (LIVE): positive — Praises the surprise reveal as amazing and describes the jump scare as effective. Claude Opus 5.5: positive — Commenter enjoyed the scares and called the surprise reveal amazing, praising the film's moments. Claude Sonnet 5.5: positive — Commenter says the jump scare got them and calls the following reveal amazing, expressing enjoyment of the film's scenes. Gemini 4 Argon: positive — The commenter praises a jump scare and calls a subsequent surprise reveal 'amazing'. Grok 4.7: positive — Praises the jump scare and surprise reveal as amazing. GPT-6 Luna (cheap): positive — They praise the effective jump scare and call the surprise reveal amazing. Claude Haiku 5.5 (cheap): positive — Commenter praises the jump scares and reveal as 'amazing', indicating enjoyment of the film's scares. Gemini 3.5 Flash-Lite (cheap): positive — The commenter praises the jump scares and calls the reveal amazing. | 0 | + | + | + | + | + | + | + | + | + | + |
| thread #29 w 0.183 The one that got me was the fakeout where he's focusing on the farm wife running towards the house through the snow when the husband comes running out of the closet. I flinched so bad. Lol VADER (old v1): negative — VADER -0.343 siebert+opinion (previous live): negative — P(op)=0.98 P(pos)=0.00 Local NLI stance (deberta-v3-large): positive — NLI p=0.40 GPT-6.1 Sol (LIVE): no_verdict — Describes a jump scare and their reaction without judging the film's quality or recommending it. Claude Opus 5.5: positive — Recounts an effective jump scare that made them flinch; enjoyment implied, no criticism. Claude Sonnet 5.5: no_verdict — Describes a jump-scare moment and their reaction, but gives no explicit judgment of the film; the scene description doesn't clearly match Resident Evil either. Gemini 4 Argon: positive — The commenter praises an effective jump scare in the film. Grok 4.7: positive — Jump scare landed and they laughed about flinching; scare worked. GPT-6 Luna (cheap): no_verdict — Describes a jump scare and personal reaction without judging the film overall. Claude Haiku 5.5 (cheap): positive — Commenter describes a jump scare that got them and laughs about it, implying enjoyment of the scene. Gemini 3.5 Flash-Lite (cheap): positive — Praises the effective jump scare and visceral reaction. | − | − | + | 0 | + | 0 | + | + | 0 | + | + |
| thread #30 w 0.18 Literally only time I actually heard people scream in the cinema. Wasn't me. For sure. VADER (old v1): negative — VADER -0.241 siebert+opinion (previous live): negative — P(op)=0.42 P(pos)=0.00 Local NLI stance (deberta-v3-large): negative — NLI p=0.33 GPT-6.1 Sol (LIVE): no_verdict — Describes audience screams and jokes about being scared, without judging the film’s quality or recommending it. Claude Opus 5.5: positive — Fondly recalls the film being uniquely scary enough to make audiences scream, implying effective horror. Claude Sonnet 5.5: no_verdict — Anecdote about audience screaming in the cinema and a joke; implies it was scary but gives no clear judgement of the film's quality. Gemini 4 Argon: positive — Describing the horror movie as genuinely scary enough to make cinema audiences scream is praise for the genre. Grok 4.7: positive — Implies the film's scares worked; joking they screamed too. GPT-6 Luna (cheap): no_verdict — They recount an audience reaction but give no clear judgment of the film. Claude Haiku 5.5 (cheap): no_verdict — Describes audience screaming in the cinema with a joke; no explicit judgement of the film's quality or recommendation. Gemini 3.5 Flash-Lite (cheap): no_verdict — Anecdote about cinema reaction without expressing personal opinion on the film. | − | − | − | 0 | + | 0 | + | + | 0 | 0 | 0 |
| thread #31 w 0.177 I didn’t expect to see Zach Cherry get absolutely destroyed by a desk in this but it got a good laugh out of me. VADER (old v1): positive — VADER 0.818 siebert+opinion (previous live): neutral — P(op)=0.31 P(pos)=0.99 Local NLI stance (deberta-v3-large): no_verdict — NLI p=0.44 GPT-6.1 Sol (LIVE): positive — Praises an unexpected scene for getting a good laugh out of them. Claude Opus 5.5: no_verdict — Amused reaction to one scene, possibly laughing at it; no judgement of the film overall. Claude Sonnet 5.5: no_verdict — Comments on a funny death scene they enjoyed, but gives no overall judgement of the film or recommendation. Gemini 4 Argon: positive — The commenter enjoyed a specific scene that gave them a good laugh. Grok 4.7: no_verdict — Joke about one funny scene, no overall film judgement. GPT-6 Luna (cheap): positive — The desk gag made them laugh, expressing enjoyment of the film. Claude Haiku 5.5 (cheap): positive — Commenter enjoyed a scene, saying it got a good laugh out of them, implying a favorable view of the film. Gemini 3.5 Flash-Lite (cheap): no_verdict — Comment focuses on a specific funny scene without giving a verdict on the film overall. | + | 0 | 0 | + | 0 | 0 | + | 0 | + | + | 0 |
| thread #32 w 0.174 I barely watched the Odyssey this week, so it was funny seeing another Antinous in this movie. Just him telling other people to do the things they aren't brave enough to do themselves. VADER (old v1): positive — VADER 0.1 siebert+opinion (previous live): negative — P(op)=0.42 P(pos)=0.00 Local NLI stance (deberta-v3-large): no_verdict — NLI p=0.79 GPT-6.1 Sol (LIVE): no_verdict — Compares a character to Antinous and discusses his behavior without judging the film. Claude Opus 5.5: no_verdict — Compares a character to Antinous from the Odyssey; offers no judgement of the film's quality. Claude Sonnet 5.5: no_verdict — Joking comparison of a character to Antinous from the Odyssey; describes a character without giving any judgment of the film itself. Gemini 4 Argon: no_verdict — The comment makes a character comparison to another work without expressing an opinion on the film's quality. Grok 4.7: no_verdict — Character comparison joke; no judgement of the film. GPT-6 Luna (cheap): no_verdict — The comment makes a character comparison but gives no opinion on the film. Claude Haiku 5.5 (cheap): no_verdict — Comment makes a joke about a character and references another film; no judgement of this film's quality or recommendation. Gemini 3.5 Flash-Lite (cheap): no_verdict — The commenter makes a casual observation comparing a character to The Odyssey without judging the film. | + | − | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| thread #33 w 0.171 The on-going padlock gag was great VADER (old v1): positive — VADER 0.402 siebert+opinion (previous live): positive — P(op)=0.58 P(pos)=1.00 Local NLI stance (deberta-v3-large): no_verdict — NLI p=0.53 GPT-6.1 Sol (LIVE): positive — Praises the film's recurring padlock gag as great. Claude Opus 5.5: positive — Praises a specific running gag in the film as great, a positive judgement of the film. Claude Sonnet 5.5: positive — Commenter praises a running gag in the film as great, which is a favorable judgment of part of the film. Gemini 4 Argon: positive — The commenter directly praises a comedic element of the film ('was great') with no accompanying criticism. Grok 4.7: positive — Praises the ongoing padlock gag as great. GPT-6 Luna (cheap): positive — They praise the film’s recurring padlock gag. Claude Haiku 5.5 (cheap): positive — Calls a specific gag in the film 'great', an explicit praise of the movie's content. Gemini 3.5 Flash-Lite (cheap): positive — Praising a specific aspect of the film as 'great'. | + | + | 0 | + | + | + | + | + | + | + | + |
| thread #34 w 0.169 Great callback to the games. Loved how many of the props were from the game too. Was really excited the first time I saw the farm gate and the big ass key. VADER (old v1): positive — VADER 0.802 siebert+opinion (previous live): neutral — P(op)=0.23 P(pos)=1.00 Local NLI stance (deberta-v3-large): no_verdict — NLI p=0.60 GPT-6.1 Sol (LIVE): positive — Praises the game's callbacks and props, expressing love and excitement for their inclusion. Claude Opus 5.5: positive — Enthusiastic praise for the game callbacks and props; clearly enjoyed the film. Claude Sonnet 5.5: positive — Commenter praises the film's callbacks to the games and says they loved the game props and were excited seeing them. Gemini 4 Argon: positive — The commenter praises the callbacks and props from the games, expressing love and excitement for those elements. Grok 4.7: positive — Praises game callbacks, props, and specific exciting moments. GPT-6 Luna (cheap): positive — They praise the game callbacks and props, saying they loved them and were excited to see them. Claude Haiku 5.5 (cheap): positive — Commenter praises the film's callbacks to the games and says they were excited by its props, showing enjoyment. Gemini 3.5 Flash-Lite (cheap): positive — The commenter expresses clear enjoyment and appreciation for the game callbacks and props. | + | 0 | 0 | + | + | + | + | + | + | + | + |
| thread #35 w 0.167 The first aid spray in the barn was the first Easter egg I saw. VADER (old v1): neutral — VADER 0.0 siebert+opinion (previous live): neutral — P(op)=0.12 P(pos)=1.00 Local NLI stance (deberta-v3-large): no_verdict — NLI p=0.54 GPT-6.1 Sol (LIVE): no_verdict — Identifies an Easter egg without expressing an opinion on the film. Claude Opus 5.5: no_verdict — Points out an Easter egg referencing the games without judging the film's quality. Claude Sonnet 5.5: no_verdict — Only points out an Easter egg (first aid spray in the barn) without judging the film's quality. Gemini 4 Argon: no_verdict — The comment simply points out an Easter egg without expressing an opinion on the film's quality. Grok 4.7: no_verdict — Notes an Easter egg without judging the film. GPT-6 Luna (cheap): no_verdict — Mentions spotting an Easter egg without expressing an opinion about the film. Claude Haiku 5.5 (cheap): no_verdict — Comment points out a detail (an Easter egg) without any judgement of whether the film is good or recommended. Gemini 3.5 Flash-Lite (cheap): no_verdict — The commenter only points out an Easter egg without expressing any judgment on the film itself. | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| thread #36 w 0.164 When he was walking past the cars in the tunnel, one of the vehicles had a giant green plant in the back seat. VADER (old v1): neutral — VADER 0.0 siebert+opinion (previous live): neutral — P(op)=0.06 P(pos)=0.00 Local NLI stance (deberta-v3-large): no_verdict — NLI p=0.72 GPT-6.1 Sol (LIVE): no_verdict — The comment notes a visual detail in a scene without evaluating the film. Claude Opus 5.5: no_verdict — Trivia observation about a background detail; no judgement of the film. Claude Sonnet 5.5: no_verdict — Commenter only points out a scene detail (a plant in a car) without judging the film. Gemini 4 Argon: no_verdict — The comment merely points out a background detail or Easter egg without expressing an opinion on the film. Grok 4.7: no_verdict — Trivia observation about a background detail, no film judgement. GPT-6 Luna (cheap): no_verdict — Points out a visual detail without expressing an opinion about the film. Claude Haiku 5.5 (cheap): no_verdict — Describes a specific scene detail with no judgement of the film's quality or whether they'd recommend it. Gemini 3.5 Flash-Lite (cheap): no_verdict — The comment only mentions a minor plot detail/trivia without expressing any opinion on the film. | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| thread #37 w 0.162 "Are we gonna abandon her???" "It's what she would've wanted" Paul Walter Hauser might have had like 5 minutes of screentime, but he killed it. VADER (old v1): negative — VADER -0.188 siebert+opinion (previous live): negative — P(op)=0.98 P(pos)=0.00 Local NLI stance (deberta-v3-large): mixed — NLI p=0.46 GPT-6.1 Sol (LIVE): positive — Praises Paul Walter Hauser’s performance, saying he "killed it" despite limited screen time. Claude Opus 5.5: positive — Quotes a funny moment and praises Paul Walter Hauser's brief performance as excellent. Claude Sonnet 5.5: positive — Praises Paul Walter Hauser's performance ('he killed it') despite brief screentime; this is an approving comment about the film's content. Gemini 4 Argon: positive — The commenter quotes a funny line and praises Paul Walter Hauser's brief performance in the film. Grok 4.7: no_verdict — Quotes a scene and praises one actor, no film verdict. GPT-6 Luna (cheap): no_verdict — Praises Paul Walter Hauser’s performance, but gives no overall judgment of the film. Claude Haiku 5.5 (cheap): positive — Praises actor Paul Walter Hauser's performance as having 'killed it' and quotes a funny line; approving tone toward the film's content. Gemini 3.5 Flash-Lite (cheap): positive — The commenter praises the actor's performance, stating he 'killed it'. | − | − | 0 | + | + | + | + | 0 | 0 | + | + |
| thread #38 w 0.16 "MAXINE? DRIVE" "Shut the fuck up" VADER (old v1): negative — VADER -0.542 siebert+opinion (previous live): negative — P(op)=0.96 P(pos)=0.00 Local NLI stance (deberta-v3-large): negative — NLI p=0.58 GPT-6.1 Sol (LIVE): no_verdict — The comment quotes dialogue without expressing a judgment of the film. Claude Opus 5.5: no_verdict — Just quotes dialogue from the film without expressing any judgement of it. Claude Sonnet 5.5: no_verdict — Just quotes dialogue from the film with no judgement of the movie itself. Gemini 4 Argon: no_verdict — The comment only provides quotes without expressing any opinion or judgment on the film. Grok 4.7: no_verdict — Just quotes movie dialogue, no opinion on the film. GPT-6 Luna (cheap): no_verdict — The comment is a quoted exchange with no opinion about the film. Claude Haiku 5.5 (cheap): no_verdict — Only quotes movie dialogue; no judgement of the film is expressed. Gemini 3.5 Flash-Lite (cheap): no_verdict — Quoting lines from the film without expressing any judgment or recommendation. | − | − | − | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| thread #39 w 0.158 Everything with Maxine was funny. Shows up and is fucking bad ass, starts blasting zombies like it's nothing. Cut to Abrams and Hauser, "fuck she's cool. Wait where's she going? No dont go in there!!!... think she's coming back?" VADER (old v1): positive — VADER 0.842 siebert+opinion (previous live): positive — P(op)=1.00 P(pos)=0.97 Local NLI stance (deberta-v3-large): positive — NLI p=0.65 GPT-6.1 Sol (LIVE): positive — Praises Maxine’s scenes as funny and badass, expressing enjoyment of the film’s action and humor. Claude Opus 5.5: positive — Enjoys Maxine's scenes, calling them funny and badass; affectionate recounting suggests liking the film. Claude Sonnet 5.5: positive — Commenter enjoys the Maxine scenes, calling them funny and her badass, which signals enjoyment of the film, though only part of it is discussed. Gemini 4 Argon: positive — The commenter praises a specific character's scenes as funny and badass, showing clear enjoyment of the film. Grok 4.7: positive — Praises Maxine scenes as funny, badass, and cool. GPT-6 Luna (cheap): positive — They praise Maxine’s scenes as funny and cool, expressing clear enjoyment of the film’s content. Claude Haiku 5.5 (cheap): positive — Praises Maxine as badass and funny, enjoying the scenes; no criticism of the film expressed. Gemini 3.5 Flash-Lite (cheap): positive — Praises a character and scene as funny and bad ass, indicating enjoyment. | + | + | + | + | + | + | + | + | + | + | + |
| thread #40 w 0.156 If this is the version with LESS jokes, what did the original possibly look like? VADER (old v1): positive — VADER 0.738 siebert+opinion (previous live): neutral — P(op)=0.34 P(pos)=0.00 Local NLI stance (deberta-v3-large): negative — NLI p=0.46 GPT-6.1 Sol (LIVE): negative — Implies the film has far too many jokes, criticizing its excessive humor. Claude Opus 5.5: no_verdict — Wry remark about the film's joke count and an earlier version; no clear judgement of the film's quality. Claude Sonnet 5.5: negative — Sarcastic quip implies the film is already full of (bad) jokes; mocking tone suggests a negative view, though no explicit verdict. Gemini 4 Argon: no_verdict — The comment makes an observation about the amount of humor in the film without giving a clear overall verdict. Grok 4.7: negative — Sarcastic jab implying the film is overloaded with jokes. GPT-6 Luna (cheap): negative — The rhetorical question implies this version already has too many jokes, criticizing the film's tone. Claude Haiku 5.5 (cheap): no_verdict — Comment asks a speculative question about the original cut's jokes; no judgement of the film's quality expressed. Gemini 3.5 Flash-Lite (cheap): no_verdict — Comment asks a question about the film's tone without expressing a clear verdict. | + | 0 | − | − | 0 | − | 0 | − | − | 0 | 0 |
| thread #41 w 0.154 I assume Yakity Sax was involved. VADER (old v1): neutral — VADER 0.0 siebert+opinion (previous live): neutral — P(op)=0.09 P(pos)=0.14 Local NLI stance (deberta-v3-large): no_verdict — NLI p=0.36 GPT-6.1 Sol (LIVE): no_verdict — A joke referencing music, without a clear judgment of the film. Claude Opus 5.5: no_verdict — Joke referencing Yakety Sax comedic chase music; no clear judgement of the film itself. Claude Sonnet 5.5: no_verdict — A joke about comedic music (Yakity Sax) with no explicit judgment of the film's quality or recommendation. Gemini 4 Argon: no_verdict — The comment is just a brief joke without any judgment or opinion on the film itself. Grok 4.7: no_verdict — Joke reference only; no clear judgement of the film. GPT-6 Luna (cheap): no_verdict — A joke referencing chase music, with no clear judgment of the film. Claude Haiku 5.5 (cheap): no_verdict — The comment is a joke/trivia reference to the Yakity Sax theme, with no judgement of the film's quality. Gemini 3.5 Flash-Lite (cheap): no_verdict — The comment is a joke/reference and does not express a judgment on the film. | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| thread #42 w 0.152 Funniest screaming since Ryan Gosling in Nice Guys VADER (old v1): positive — VADER 0.586 siebert+opinion (previous live): positive — P(op)=0.59 P(pos)=1.00 Local NLI stance (deberta-v3-large): mixed — NLI p=0.45 GPT-6.1 Sol (LIVE): positive — Praises the screaming as exceptionally funny, favorably comparing it to Ryan Gosling in The Nice Guys. Claude Opus 5.5: no_verdict — A joke comparing a character's screaming to The Nice Guys; it gives no clear judgement of the film. Claude Sonnet 5.5: no_verdict — A joking comparison about a screaming scene; it doesn't clearly state whether they liked or recommend the film overall. Gemini 4 Argon: no_verdict — The comment makes a joke/observation about screaming in the film without giving an overall verdict on the movie itself. Grok 4.7: no_verdict — Joke comparing screaming; no overall film judgment. GPT-6 Luna (cheap): positive — Calls the screaming funny, expressing praise for an element of the film. Claude Haiku 5.5 (cheap): no_verdict — Joke comparing screaming to Ryan Gosling in Nice Guys; no clear judgement of Resident Evil itself. Gemini 3.5 Flash-Lite (cheap): positive — Comparing the screaming to a well-liked comedy moment implies enjoyment of the scene. | + | + | 0 | + | 0 | 0 | 0 | 0 | + | 0 | + |
| thread #43 w 0.151 Bryan's screams when he running through the tall buildings as the creatures fall off them straight up feels like something you see in WKUK lmfao VADER (old v1): positive — VADER 0.691 siebert+opinion (previous live): negative — P(op)=0.97 P(pos)=0.01 Local NLI stance (deberta-v3-large): negative — NLI p=0.41 GPT-6.1 Sol (LIVE): no_verdict — Compares a scene to WKUK humorously without expressing an overall opinion of the film. Claude Opus 5.5: no_verdict — A joke comparing one scene to sketch comedy, with no clear overall judgement of the film. Claude Sonnet 5.5: no_verdict — Jokey remark about a scene resembling a WKUK sketch; amused but gives no clear overall judgment or recommendation of the film. Gemini 4 Argon: no_verdict — The comment makes a joke about a specific character's scream without expressing an overall opinion on the movie. Grok 4.7: no_verdict — Joke comparing a scene to comedy sketches; no film judgment. GPT-6 Luna (cheap): no_verdict — Compares a scene to a comedy sketch but gives no judgment of the film. Claude Haiku 5.5 (cheap): no_verdict — Compares a scene to another film for humor; no judgement of this film's quality or recommendation. Gemini 3.5 Flash-Lite (cheap): no_verdict — The comment makes a joke about a specific scene compared to a comedy sketch without judging the film overall. | + | − | − | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| thread #44 w 0.149 Once Bryan recorded his message his wife would never hear, I knew what Creggar was going for. This is essentially the off-screen minor story you read about in documents you find while wandering. There's always bit characters recording their journey until they end with a tragic… VADER (old v1): positive — VADER 0.459 siebert+opinion (previous live): neutral — P(op)=0.30 P(pos)=1.00 Local NLI stance (deberta-v3-large): no_verdict — NLI p=0.80 GPT-6.1 Sol (LIVE): positive — Calls the storytelling brilliant and praises keeping Resident Evil terrifying while avoiding overpowered protagonists. Claude Opus 5.5: positive — Calls the film's approach brilliant, keeping Resident Evil terrifying throughout the runtime; analysis is appreciative. Claude Sonnet 5.5: positive — Commenter praises the film's approach as 'brilliant' for capturing RE's game feel and keeping it terrifying throughout. Gemini 4 Argon: positive — Calls the film's approach 'brilliant' and praises it for staying 'terrifying throughout the run-time'. Grok 4.7: positive — Praises the film's approach as brilliant and terrifying throughout. GPT-6 Luna (cheap): positive — Calls the storytelling approach brilliant and praises how it keeps the film terrifying. Claude Haiku 5.5 (cheap): positive — Commenter praises the film's storytelling as brilliant and says it keeps Resident Evil terrifying throughout the runtime. Gemini 3.5 Flash-Lite (cheap): positive — Praises the film's approach as a brilliant way to keep it terrifying. | + | 0 | 0 | + | + | + | + | + | + | + | + |
| thread #45 w 0.147 Yep, my first thought as soon as he pulled up the audio recorder app was “ohhhh he’s an NPC”. Felt like a cassette you’d find in Bioshock. VADER (old v1): positive — VADER 0.572 siebert+opinion (previous live): negative — P(op)=0.42 P(pos)=0.00 Local NLI stance (deberta-v3-large): no_verdict — NLI p=0.61 GPT-6.1 Sol (LIVE): no_verdict — Compares a character moment to video game storytelling without evaluating the film. Claude Opus 5.5: no_verdict — Observation comparing a scene to video game tropes; no judgement of the film's quality. Claude Sonnet 5.5: no_verdict — Describes a scene and compares it to a game trope (NPC, Bioshock cassette) without judging the film's quality or recommending it. Gemini 4 Argon: no_verdict — The comment makes a joke/observation comparing a character moment to a video game trope without judging the film. Grok 4.7: no_verdict — Scene joke comparing a character to game tropes, no film judgement. GPT-6 Luna (cheap): no_verdict — They compare a scene to a video game trope but don't express an overall opinion of the film. Claude Haiku 5.5 (cheap): no_verdict — Comments on a scene's resemblance to Bioshock lore without judging the film itself; no verdict expressed. Gemini 3.5 Flash-Lite (cheap): no_verdict — The commenter makes a joke/comparison about a character's action without judging the film itself. | + | − | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| thread #46 w 0.146 great detail I didn't think of. you'll definitely pick up a phone while searching for supplies, listen to 2 seconds and move on VADER (old v1): positive — VADER 0.778 siebert+opinion (previous live): neutral — P(op)=0.33 P(pos)=0.01 Local NLI stance (deberta-v3-large): no_verdict — NLI p=0.55 GPT-6.1 Sol (LIVE): no_verdict — Discusses the plausibility of a specific detail without expressing an overall opinion of the film. Claude Opus 5.5: no_verdict — Replies to a plot detail with a joke about realistic behavior. No judgement of the film. Claude Sonnet 5.5: no_verdict — Commenter discusses a detail about phone messages in the game/film without giving an overall judgement of the film itself. Gemini 4 Argon: no_verdict — Discusses a minor plot/realism detail without expressing an overall opinion on the film. Grok 4.7: no_verdict — Praises a gameplay detail, not the film overall. GPT-6 Luna (cheap): positive — They praise a detail in the film as thoughtful and realistic. Claude Haiku 5.5 (cheap): no_verdict — Comment discusses a plot/realism detail about phones and supplies without giving any judgement of the film itself. Gemini 3.5 Flash-Lite (cheap): no_verdict — Discusses a gameplay detail or joke rather than giving a verdict on the film itself. | + | 0 | 0 | 0 | 0 | 0 | 0 | 0 | + | 0 | 0 |
| thread #47 w 0.144 My sewer... My rats...My Racoon City VADER (old v1): neutral — VADER 0.0 siebert+opinion (previous live): positive — P(op)=0.43 P(pos)=1.00 Local NLI stance (deberta-v3-large): negative — NLI p=0.43 GPT-6.1 Sol (LIVE): no_verdict — A joke or quote referencing the setting, without any judgment of the film. Claude Opus 5.5: no_verdict — A joke or quote riff that makes no judgement of the film's quality. Claude Sonnet 5.5: no_verdict — A joking, quote-style remark with no judgement of the film's quality or any recommendation. Gemini 4 Argon: no_verdict — The comment is just a joke/quote parodying Dune with Resident Evil elements, offering no opinion on the movie. Grok 4.7: no_verdict — Quote or joke with no judgment of the film. GPT-6 Luna (cheap): no_verdict — A joking reference to the movie's setting, with no opinion about the film. Claude Haiku 5.5 (cheap): no_verdict — Playful quote from the film with no clear judgement of whether the commenter liked it. Gemini 3.5 Flash-Lite (cheap): no_verdict — Quoting a line from the movie, no judgment expressed. | 0 | + | − | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| thread #49 w 0.141 "He's FUCKING INFECTED!" "No, no, I'm fine. Just a lot of stairs." This entire part had me wheezing. VADER (old v1): negative — VADER -0.665 siebert+opinion (previous live): negative — P(op)=0.97 P(pos)=0.00 Local NLI stance (deberta-v3-large): negative — NLI p=0.40 GPT-6.1 Sol (LIVE): positive — The commenter praises the scene's humor, saying it had them wheezing with laughter. Claude Opus 5.5: positive — Quotes a scene and says it had them laughing hard, showing clear enjoyment of the film. Claude Sonnet 5.5: positive — Commenter enjoyed a scene so much they were wheezing with laughter, signaling enjoyment of the film, though no overall verdict is given. Gemini 4 Argon: positive — The commenter quotes a scene and expresses strong enjoyment by saying it had them 'wheezing' with laughter. Grok 4.7: positive — Found a scene hilarious; amused reaction is praise of the film. GPT-6 Luna (cheap): positive — They say the scene had them wheezing, expressing enjoyment of the film's humor. Claude Haiku 5.5 (cheap): positive — The commenter found a scene hilarious ('had me wheezing'), signaling enjoyment of this part of the film. Gemini 3.5 Flash-Lite (cheap): positive — The commenter found a funny scene highly amusing, indicating they enjoyed the movie. | − | − | − | + | + | + | + | + | + | + | + |
| thread #50 w 0.14 What does he say, "Think I could get a little bit of that?" As he's rapidly transforming into the thing from hell. Lol VADER (old v1): negative — VADER -0.421 siebert+opinion (previous live): neutral — P(op)=0.06 P(pos)=0.00 Local NLI stance (deberta-v3-large): positive — NLI p=0.38 GPT-6.1 Sol (LIVE): no_verdict — Quotes and jokes about a character's transformation without judging the film. Claude Opus 5.5: no_verdict — Quotes and jokes about a scene without judging the film's quality. Claude Sonnet 5.5: no_verdict — Recalls a scene and laughs at it, but gives no clear judgement on whether the film is good or bad. Gemini 4 Argon: no_verdict — The comment simply quotes and describes a funny moment from the movie without giving an overall verdict. Grok 4.7: no_verdict — Quotes a line and jokes about a scene; no film judgement. GPT-6 Luna (cheap): no_verdict — Jokes about a character’s line during a transformation scene without judging the film. Claude Haiku 5.5 (cheap): no_verdict — Quotes a line and jokes about a scene; no judgement of the film's quality or whether they'd recommend it. Gemini 3.5 Flash-Lite (cheap): no_verdict — Quoting a funny line and discussing a scene without evaluating the film. | − | 0 | + | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| thread #51 w 0.139 Before I start: The RV scene is the best scene in the movie. We finally got a GREAT Resident Evil movie. I don't know how Cregger captured the feeling of playing these games with such clarity. To balance horror with the absurdity of a situation or to show the ridiculousness of… VADER (old v1): positive — VADER 0.873 siebert+opinion (previous live): positive — P(op)=0.96 P(pos)=1.00 Local NLI stance (deberta-v3-large): positive — NLI p=0.95 GPT-6.1 Sol (LIVE): positive — Calls it a great, phenomenal adaptation that captures the games’ feeling and balances horror with absurdity. Claude Opus 5.5: positive — Calls it a great Resident Evil movie, phenomenal, praising how it captures the games' feeling. Claude Sonnet 5.5: positive — Calls it a GREAT Resident Evil movie, praises Cregger's capturing of the games' feel, and ends with 'Phenomenal.' Gemini 4 Argon: positive — The commenter explicitly calls it a "GREAT Resident Evil movie" and "Phenomenal." Grok 4.7: positive — Calls it a great, phenomenal Resident Evil movie that captures the games. GPT-6 Luna (cheap): positive — Calls it a great, phenomenal Resident Evil movie and praises how it captures the games’ feeling. Claude Haiku 5.5 (cheap): positive — VADER 0.873 Gemini 3.5 Flash-Lite (cheap): positive — The commenter calls it a 'GREAT Resident Evil movie' and praises it as phenomenal. | + | + | + | + | + | + | + | + | + | + | + |
| thread #52 w 0.137 I think he captured it by making the protagonist feel less like a video game character and more like a normal person reacting while playing the games. He’s like the movie characterization of the player. VADER (old v1): positive — VADER 0.807 siebert+opinion (previous live): positive — P(op)=0.43 P(pos)=1.00 Local NLI stance (deberta-v3-large): no_verdict — NLI p=0.73 GPT-6.1 Sol (LIVE): positive — Praises the filmmaker’s protagonist characterization as successfully capturing the experience of playing the games. Claude Opus 5.5: positive — Says the director 'captured it', praising how the protagonist works as a stand-in for the player. Claude Sonnet 5.5: no_verdict — Describes how the protagonist was characterized, an analytical observation about the film's approach, without explicitly stating whether they liked it or would recommend it. Gemini 4 Argon: positive — Praises how the director captured the protagonist's characterization as a relatable proxy for the player. Grok 4.7: positive — Praises the film for capturing the player through a grounded protagonist. GPT-6 Luna (cheap): positive — They praise how the film characterizes the protagonist as a believable stand-in for the player. Claude Haiku 5.5 (cheap): no_verdict — Analyzes the protagonist's characterization as a player stand-in without stating whether they liked or recommend the film. Gemini 3.5 Flash-Lite (cheap): positive — Praises the director's successful characterization and execution. | + | + | 0 | + | + | 0 | + | + | + | 0 | + |
| thread #53 w 0.136 Austin Abrams is a star, man. It's so hard to carry a movie entirely on your back like this and he just owns it, even when he's acting against nothing. This is like Tom Hanks in Castaway if Wilson was a giant tentacle monster instead of a Volleyball VADER (old v1): negative — VADER -0.122 siebert+opinion (previous live): positive — P(op)=0.98 P(pos)=1.00 Local NLI stance (deberta-v3-large): no_verdict — NLI p=0.52 GPT-6.1 Sol (LIVE): positive — Enthusiastically praises Abrams’s lead performance and ability to carry the movie, with an approving Castaway comparison. Claude Opus 5.5: positive — Strong praise for the lead's performance carrying the movie, compared favorably to Hanks in Cast Away. Claude Sonnet 5.5: positive — Strongly praises Austin Abrams' performance as carrying the film and compares it favorably to Castaway, implying approval of the movie. Gemini 4 Argon: positive — Strong praise for Austin Abrams' performance carrying the film, with no criticism. Grok 4.7: positive — Praises the lead for carrying and owning the movie. GPT-6 Luna (cheap): positive — Enthusiastically praises Abrams for carrying the movie, with only a mild jab at what he has to act against. Claude Haiku 5.5 (cheap): positive — Praises the lead actor as a star who 'owns it' and compares the film favorably to Castaway. Gemini 3.5 Flash-Lite (cheap): positive — High praise for the lead actor's performance and the film's premise. | − | + | 0 | + | + | + | + | + | + | + | + |
| thread #54 w 0.135 Can I say something kind of weird? It was so oddly refreshing when he took off his shirt and he just looked like a normal guy. We're so conditioned to every shirtless actor having the 30 hour dehydration 6 pack. VADER (old v1): positive — VADER 0.297 siebert+opinion (previous live): neutral — P(op)=0.06 P(pos)=1.00 Local NLI stance (deberta-v3-large): mixed — NLI p=0.42 GPT-6.1 Sol (LIVE): no_verdict — Appreciates an actor's normal-looking physique without judging the film itself. Claude Opus 5.5: no_verdict — Comments on an actor's normal physique as refreshing, but gives no judgement of the film overall. Claude Sonnet 5.5: no_verdict — Comments on an actor's realistic physique being refreshing; this isn't a judgement of the film overall. Gemini 4 Argon: no_verdict — The comment only remarks on an actor's physique in one scene without expressing an overall opinion on the film. Grok 4.7: no_verdict — Observes actor's normal physique; no film judgement. GPT-6 Luna (cheap): positive — They find the film’s more naturalistic depiction of a shirtless actor refreshing. Claude Haiku 5.5 (cheap): no_verdict — Comments on an actor's physique being refreshing; no judgement of the film itself expressed. Gemini 3.5 Flash-Lite (cheap): positive — Appreciated a realistic physical depiction of an actor, indicating a favorable impression of the film. | + | 0 | 0 | 0 | 0 | 0 | 0 | 0 | + | 0 | + |
| thread #55 w 0.134 Him trying to cover himself with his shirt when he realizes he’s on camera was one of my favorite moments of the movie, and it got the biggest laugh out of my audience. It’s so relatable. VADER (old v1): positive — VADER 0.765 siebert+opinion (previous live): positive — P(op)=0.86 P(pos)=1.00 Local NLI stance (deberta-v3-large): no_verdict — NLI p=0.77 GPT-6.1 Sol (LIVE): positive — Praises a favorite moment as funny and relatable, indicating enjoyment of the film. Claude Opus 5.5: positive — Enjoys a favorite moment that got big audience laughs; implies enjoyment of the film. Claude Sonnet 5.5: positive — Commenter calls a specific scene one of their favorite moments of the movie and notes it got big laughs, implying enjoyment. Gemini 4 Argon: positive — Praises a specific humorous scene as one of their favorite moments and relatable. Grok 4.7: positive — Calls a scene one of their favorite moments and enjoyed the audience laugh. GPT-6 Luna (cheap): positive — They call the moment a favorite and praise its humor and relatability. Claude Haiku 5.5 (cheap): positive — Calls the moment one of their favorite parts and says it got the biggest laugh, a clear endorsement of the film. Gemini 3.5 Flash-Lite (cheap): positive — The commenter highlights a favorite moment and a great audience reaction, indicating they liked the film. | + | + | 0 | + | + | + | + | + | + | + | + |
| thread #56 w 0.132 He nailed the dialogue I repeat to myself when playing a resident evil game “Mother fuckin cock suckin bitch fuck what the fuck why the fuck did this god dammit fuck” VADER (old v1): negative — VADER -0.958 siebert+opinion (previous live): positive — P(op)=0.86 P(pos)=0.99 Local NLI stance (deberta-v3-large): positive — NLI p=0.38 GPT-6.1 Sol (LIVE): no_verdict — Jokes about dialogue matching their reactions while playing Resident Evil games, without judging the film itself. Claude Opus 5.5: no_verdict — Joke relating a character's profane line to playing the games; no judgement of the film's quality. Claude Sonnet 5.5: no_verdict — Joke about the character's profanity echoing the commenter's own gameplay reactions; praises the dialogue slightly but gives no clear verdict on the film. Gemini 4 Argon: no_verdict — The commenter jokes about a relatable line of dialogue without giving an overall verdict on the film itself. Grok 4.7: no_verdict — Joke about dialogue matching game frustration, no film judgement. GPT-6 Luna (cheap): positive — They praise how well he delivers the dialogue. Claude Haiku 5.5 (cheap): no_verdict — Joke about the game's dialogue and the player's frustration; no judgement of the film itself. Gemini 3.5 Flash-Lite (cheap): no_verdict — The comment is a joke quoting dialogue and does not express a clear verdict on the film. | − | + | + | 0 | 0 | 0 | 0 | 0 | + | 0 | 0 |
| thread #57 w 0.131 I love how he keeps shooting the locks. Like an answer to those gaming moments on “why cant I just shoot the lock with a freaking gun?” VADER (old v1): positive — VADER 0.713 siebert+opinion (previous live): positive — P(op)=0.56 P(pos)=1.00 Local NLI stance (deberta-v3-large): no_verdict — NLI p=0.72 GPT-6.1 Sol (LIVE): positive — Praises the film’s lock-shooting moments as a satisfying answer to a common gaming frustration. Claude Opus 5.5: positive — Expresses enjoyment of a film detail, the lock-shooting, as a clever nod to gaming frustrations. Claude Sonnet 5.5: positive — Commenter expresses affection ('I love how') for a specific film choice, a clever nod to game frustrations, implying approval of the film. Gemini 4 Argon: positive — Expresses enjoyment of a specific trope subversion and detail in the film. Grok 4.7: positive — Praises a specific film moment as a satisfying answer to game logic. GPT-6 Luna (cheap): positive — They explicitly enjoy a clever, entertaining moment in the film. Claude Haiku 5.5 (cheap): positive — VADER 0.713 Gemini 3.5 Flash-Lite (cheap): positive — Praises a specific satisfying moment in the film related to video game tropes. | + | + | 0 | + | + | + | + | + | + | + | + |
| thread #58 w 0.13 I liked how the first one was indestructible because it had a special key. (Pretty sure it had a diamond symbol too), like one of those puzzle locks that absolutely requires you to complete a puzzle. But then the one later in the movie that actually breaks is just a generic… VADER (old v1): positive — VADER 0.805 siebert+opinion (previous live): negative — P(op)=0.99 P(pos)=0.00 Local NLI stance (deberta-v3-large): no_verdict — NLI p=0.70 GPT-6.1 Sol (LIVE): positive — Expresses appreciation for the film’s distinction between a special puzzle lock and a generic, breakable padlock. Claude Opus 5.5: no_verdict — Observational detail about props/locks; 'liked' refers to a minor detail, not an overall judgement of the film. Claude Sonnet 5.5: no_verdict — Comment discusses a specific detail about locks/props in the film with mild appreciation, but gives no overall verdict on the film. Gemini 4 Argon: no_verdict — The comment merely observes a small detail about locks in the movie without judging the film overall. Grok 4.7: no_verdict — Plot detail about locks; no overall film judgement. GPT-6 Luna (cheap): mixed — They praise the special-key lock detail but criticize the later generic padlock inconsistency. Claude Haiku 5.5 (cheap): no_verdict — Discusses plot details about a lock's design and continuity; no stated judgement of the film's quality. Gemini 3.5 Flash-Lite (cheap): no_verdict — The comment discusses movie details and puzzles without giving an overall verdict on the film. | + | − | 0 | + | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| thread #59 w 0.129 The shock of seeing the little girl foot was one of the best horror moments of the year. This movie was fucking gross in all the right ways. VADER (old v1): negative — VADER -0.201 siebert+opinion (previous live): positive — P(op)=1.00 P(pos)=0.98 Local NLI stance (deberta-v3-large): positive — NLI p=0.83 GPT-6.1 Sol (LIVE): positive — Praises the shocking horror moment and explicitly enjoys the movie’s grossness. Claude Opus 5.5: positive — Praises a shocking horror moment as one of the year's best and calls the gore effective. Claude Sonnet 5.5: positive — Praises a shocking scene as one of the year's best horror moments and says the film was gross 'in all the right ways'. Gemini 4 Argon: positive — Praises a scene as one of the best horror moments and calls the movie gross 'in all the right ways'. Grok 4.7: positive — Praises a horror moment and calls the grossness exactly right. GPT-6 Luna (cheap): positive — Calls the horror moment one of the year's best and praises the movie's grossness. Claude Haiku 5.5 (cheap): positive — Commenter praises the gross, shocking horror moment and says the film was gross 'in all the right ways'. Gemini 3.5 Flash-Lite (cheap): positive — Praises the horror elements and calls the movie 'gross in all the right ways'. | − | + | + | + | + | + | + | + | + | + | + |
| thread #60 w 0.128 The foot was one thing but I was not ready for her face to open up like that. It’s like with the dog. I should had seen it coming but the movie lulled me into thinking no this was a normal dog. VADER (old v1): positive — VADER 0.258 siebert+opinion (previous live): negative — P(op)=0.45 P(pos)=0.00 Local NLI stance (deberta-v3-large): mixed — NLI p=0.53 GPT-6.1 Sol (LIVE): no_verdict — Describes surprise at horror transformations without expressing an overall opinion of the film. Claude Opus 5.5: no_verdict — Describes being surprised by gory scenes without clearly praising or criticising the film. Claude Sonnet 5.5: no_verdict — Describes shocking scenes and their reaction to them, but gives no explicit judgment of the film's quality or recommendation. Gemini 4 Argon: no_verdict — The commenter reacts to specific surprising scenes in the plot without giving an overall evaluation of the film. Grok 4.7: no_verdict — Describes surprise at body-horror scenes without judging the film. GPT-6 Luna (cheap): no_verdict — Describes being surprised by a disturbing scene, without judging the film itself. Claude Haiku 5.5 (cheap): no_verdict — Describes shocking scenes and being surprised by them, but gives no overall judgement of the film's quality or recommendation. Gemini 3.5 Flash-Lite (cheap): no_verdict — The commenter discusses specific shocking scenes without expressing an overall opinion or recommendation about the film. | + | − | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| thread #61 w 0.127 This movie was so cruel for that. I almost feel like it's why it had to be so funny, because otherwise seeing a cute dog, a sad little girl, and a bunch of babies getting ripped apart by disease would be too sad VADER (old v1): negative — VADER -0.477 siebert+opinion (previous live): negative — P(op)=1.00 P(pos)=0.00 Local NLI stance (deberta-v3-large): negative — NLI p=0.97 GPT-6.1 Sol (LIVE): no_verdict — Discusses distressing content and humor balancing its sadness, without expressing an overall judgment of the film. Claude Opus 5.5: no_verdict — Interprets the film's cruel content and humor as tonal balance without clearly judging whether they liked it. Claude Sonnet 5.5: no_verdict — Describes the film's cruelty and humor and speculates on why it's funny, but gives no clear overall judgment of liking or disliking it. Gemini 4 Argon: positive — Praises the film's humor for effectively balancing out its dark and cruel elements. Grok 4.7: no_verdict — Analyzes the film's cruel-funny tone without judging quality. GPT-6 Luna (cheap): positive — They suggest the film’s humor effectively balances its cruel and disturbing imagery. Claude Haiku 5.5 (cheap): no_verdict — Discusses the film's tone and why it uses humor, describing dark scenes without clearly judging whether they liked the film overall. Gemini 3.5 Flash-Lite (cheap): no_verdict — Comment discusses plot elements and emotional impact without explicitly judging the film's quality. | − | − | − | 0 | 0 | 0 | + | 0 | + | 0 | 0 |
| thread #62 w 0.126 Can we talk about how VADER (old v1): neutral — VADER 0.0 siebert+opinion (previous live): neutral — P(op)=0.38 P(pos)=0.00 Local NLI stance (deberta-v3-large): negative — NLI p=0.35 GPT-6.1 Sol (LIVE): no_verdict — The incomplete comment expresses no opinion about the film. Claude Opus 5.5: no_verdict — Comment is truncated and gives no judgement of the film. Claude Sonnet 5.5: no_verdict — The comment is cut off and expresses no opinion on the film. Gemini 4 Argon: no_verdict — The comment is an incomplete sentence and expresses no opinion on the movie. Grok 4.7: no_verdict — Incomplete fragment with no judgment of the film. GPT-6 Luna (cheap): no_verdict — The comment is incomplete and does not express an opinion about the film. Claude Haiku 5.5 (cheap): no_verdict — Comment is an incomplete opener with no judgement of the film expressed. Gemini 3.5 Flash-Lite (cheap): no_verdict — Incomplete sentence without any opinion on the film. | 0 | 0 | − | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| thread #63 w 0.125 Yeah what's up with that? I thought she went off to save that one lady from the zombies but after killing all of them she continued on in the dark? VADER (old v1): negative — VADER -0.697 siebert+opinion (previous live): neutral — P(op)=0.16 P(pos)=0.00 Local NLI stance (deberta-v3-large): no_verdict — NLI p=0.62 GPT-6.1 Sol (LIVE): no_verdict — Questions a character's actions and scene logic without expressing an overall judgment of the film. Claude Opus 5.5: no_verdict — Commenter asks a plot question about a scene without judging the film's quality. Claude Sonnet 5.5: no_verdict — Commenter is asking a plot question about a character's actions, without giving any judgement of the film. Gemini 4 Argon: no_verdict — The comment merely questions a specific plot point without expressing an overall opinion on the film. Grok 4.7: no_verdict — Plot confusion about a scene, no film judgement. GPT-6 Luna (cheap): no_verdict — Asks about a character's actions in the plot without judging the film. Claude Haiku 5.5 (cheap): no_verdict — Asks a plot question about what happens in the film; no judgement of the film's quality expressed. Gemini 3.5 Flash-Lite (cheap): no_verdict — The commenter is only asking a question about the plot and does not express an opinion on the film. | − | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| thread #64 w 0.124 *Ominous and dark corridor where there's very likely a murderous mutant psycho monster lying in wait" Bryan: "Hello", "HELLO" VADER (old v1): negative — VADER -0.882 siebert+opinion (previous live): neutral — P(op)=0.19 P(pos)=0.99 Local NLI stance (deberta-v3-large): no_verdict — NLI p=0.33 GPT-6.1 Sol (LIVE): no_verdict — Jokes about a character's behavior in a dangerous scene without judging the film. Claude Opus 5.5: no_verdict — Joke mocking a character's behavior in a scene; no judgement of the film itself. Claude Sonnet 5.5: no_verdict — A joke mocking a character's behavior in a horror scene; it gives no clear overall judgment of the film. Gemini 4 Argon: no_verdict — The comment makes a joke about a character's actions in a scene without expressing an overall opinion on the film. Grok 4.7: no_verdict — Joke about a scene; no judgement of the film. GPT-6 Luna (cheap): no_verdict — A joke about a character entering a dangerous corridor; it gives no opinion on the film. Claude Haiku 5.5 (cheap): no_verdict — Quotes a scene with a joke about a monster; no judgement of the film's quality or whether they'd recommend it. Gemini 3.5 Flash-Lite (cheap): no_verdict — Quoting a humorous movie trope without expressing a judgment or verdict on the film. | − | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| thread #65 w 0.123 Sewer Baron Harkonnen has entered the chat VADER (old v1): neutral — VADER 0.0 siebert+opinion (previous live): neutral — P(op)=0.28 P(pos)=0.99 Local NLI stance (deberta-v3-large): mixed — NLI p=0.29 GPT-6.1 Sol (LIVE): no_verdict — A character-reference joke without any judgment of the film. Claude Opus 5.5: no_verdict — A joke referencing another character, with no judgement of the film itself. Claude Sonnet 5.5: no_verdict — A joke referencing a character or actor resemblance (Baron Harkonnen) with no judgement of the film itself. Gemini 4 Argon: no_verdict — The comment is a joke about a character or creature without expressing an opinion on the movie. Grok 4.7: no_verdict — Joke reference only; no judgement of the film. GPT-6 Luna (cheap): no_verdict — A joke comparing a character to Baron Harkonnen, with no judgment of the film. Claude Haiku 5.5 (cheap): no_verdict — A joke referencing a character from another franchise; no judgement of this film's quality. Gemini 3.5 Flash-Lite (cheap): no_verdict — The comment is a joke/reference comparing a character to Baron Harkonnen without judging the film. | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| thread #66 w 0.122 The way the entire theater groaned (complimentary) that he has to take the stairs all the way up to the top is now a core memory this year in theaters. VADER (old v1): positive — VADER 0.572 siebert+opinion (previous live): negative — P(op)=0.59 P(pos)=0.00 Local NLI stance (deberta-v3-large): negative — NLI p=0.44 GPT-6.1 Sol (LIVE): positive — Describes the scene's audience reaction as complimentary and a memorable theatrical experience. Claude Opus 5.5: positive — Describes a fond theater moment as a 'core memory', and marks the groan as complimentary, implying enjoyment. Claude Sonnet 5.5: positive — Commenter fondly recalls the audience's reaction as a core memory of the year in theaters, implying an enjoyable viewing experience. Gemini 4 Argon: positive — The commenter fondly praises a memorable theater experience and explicitly labels the audience's reaction as complimentary. Grok 4.7: positive — Calls the theater groan complimentary and a core memory, implying enjoyment. GPT-6 Luna (cheap): positive — They fondly describe a memorable, entertaining theater reaction to the scene. Claude Haiku 5.5 (cheap): positive — Commenter describes a theater moment as a 'core memory' with 'complimentary' groaning, implying enjoyment of the film's fun, over-the-top scene. Gemini 3.5 Flash-Lite (cheap): positive — The commenter shares a memorable, highly positive theater experience related to watching the film. | + | − | − | + | + | + | + | + | + | + | + |
| thread #67 w 0.121 In my head canon cregger is also a final fantasy 7 fan and this was a way for him to pay homage to another video game he loves 😅 VADER (old v1): positive — VADER 0.796 siebert+opinion (previous live): neutral — P(op)=0.10 P(pos)=1.00 Local NLI stance (deberta-v3-large): no_verdict — NLI p=0.50 GPT-6.1 Sol (LIVE): no_verdict — Speculates about the filmmaker’s inspiration and video game homage without judging the film. Claude Opus 5.5: no_verdict — Playful head-canon about the director's influences; no judgement of the film's quality. Claude Sonnet 5.5: no_verdict — Speculative, joking head canon about the director's influences; no judgment of the film's quality or recommendation. Gemini 4 Argon: no_verdict — The comment shares a fun theory about the director without expressing any opinion on the film's quality. Grok 4.7: no_verdict — Director homage joke; no judgement of the film. GPT-6 Luna (cheap): no_verdict — Speculates about the creator’s inspirations without judging the film. Claude Haiku 5.5 (cheap): no_verdict — Speculative headcanon about the director's influences; no judgement of the film's quality or whether they liked it. Gemini 3.5 Flash-Lite (cheap): no_verdict — The comment is a joke about head canon and video game homages, not a judgment of the film. | + | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| thread #68 w 0.12 This is literally me playing Resident Evil games. Missing shots✓ Saying Fuck✓ Pissed at locks✓ VADER (old v1): negative — VADER -0.875 siebert+opinion (previous live): negative — P(op)=0.54 P(pos)=0.00 Local NLI stance (deberta-v3-large): negative — NLI p=0.50 GPT-6.1 Sol (LIVE): no_verdict — Jokes about similarities to their gameplay experience without judging the film. Claude Opus 5.5: no_verdict — Joke relating the characters to the commenter's own gameplay; no judgement of the film. Claude Sonnet 5.5: no_verdict — A joke relating the film's scene to the commenter's own gameplay experience; no judgment of the film's quality. Gemini 4 Argon: no_verdict — The comment makes a joke comparing the movie to playing the video games without judging the film's quality. Grok 4.7: no_verdict — Joke comparing the film to playing the games, no film judgment. GPT-6 Luna (cheap): no_verdict — A joke about playing the games, with no judgment of the film. Claude Haiku 5.5 (cheap): no_verdict — The comment jokingly compares the speaker's gameplay habits to the game, with no judgement of the film itself. Gemini 3.5 Flash-Lite (cheap): positive — Relates the movie to playing the games with humor, implying a fun, enjoyable experience. | − | − | − | 0 | 0 | 0 | 0 | 0 | 0 | 0 | + |
| thread #69 w 0.12 Saying hell yeah to the shotgun on the wall VADER (old v1): negative — VADER -0.527 siebert+opinion (previous live): positive — P(op)=0.76 P(pos)=1.00 Local NLI stance (deberta-v3-large): positive — NLI p=0.52 GPT-6.1 Sol (LIVE): no_verdict — Expresses excitement about a shotgun in a scene, without judging the film overall. Claude Opus 5.5: no_verdict — Enthusiastic reaction to one scene detail, not a judgement of the film overall. Claude Sonnet 5.5: no_verdict — Enthusiastic reaction to a specific scene/prop, but it doesn't clearly state an overall verdict or recommendation of the film. Gemini 4 Argon: no_verdict — Expressing excitement over a specific easter egg or prop without judging the overall film. Grok 4.7: positive — Enthusiastic hell-yeah reaction to a film moment shows enjoyment. GPT-6 Luna (cheap): no_verdict — They react enthusiastically to a shotgun on the wall but don't judge the film itself. Claude Haiku 5.5 (cheap): no_verdict — Comment is an enthusiastic reference to a shotgun prop, not a clear judgement of the film's quality. Gemini 3.5 Flash-Lite (cheap): positive — Praising a specific iconic scene indicates enjoyment of the film. | − | + | + | 0 | 0 | 0 | 0 | + | 0 | 0 | + |
| thread #70 w 0.119 That was a lot of fun. I liked that whatever the virus is in this world seems to be a mix of Las Plagas from RE 4 with a sprinkle of The Thing. The elevator scene was fun cause I legit forgot about the acid throw up. My only issue is how abrupt the end is. Unlike the recent… VADER (old v1): positive — VADER 0.956 siebert+opinion (previous live): positive — P(op)=0.98 P(pos)=1.00 Local NLI stance (deberta-v3-large): mixed — NLI p=0.57 GPT-6.1 Sol (LIVE): positive — Repeatedly calls it fun, praises the virus and elevator scene, and considers rewatching despite the abrupt ending. Claude Opus 5.5: positive — Called it a lot of fun and may rewatch; only complaint is the abrupt ending. Claude Sonnet 5.5: positive — Calls it 'a lot of fun' repeatedly and may see it again; only complaint is the abrupt ending. Gemini 4 Argon: positive — The commenter explicitly states the movie was a lot of fun and wants to see it again despite an abrupt ending. Grok 4.7: positive — Calls it fun and may rewatch, despite abrupt ending. GPT-6 Luna (cheap): positive — They call it fun, say they liked it, and may rewatch; the abrupt ending is a minor criticism. Claude Haiku 5.5 (cheap): positive — VADER 0.956 Gemini 3.5 Flash-Lite (cheap): positive — The commenter calls the movie 'a lot of fun' multiple times and considers seeing it again. | + | + | 0 | + | + | + | + | + | + | + | + |
| thread #71 w 0.118 Him putting the throw up zombie in a chokehold was fucking hilarious. VADER (old v1): positive — VADER 0.458 siebert+opinion (previous live): positive — P(op)=0.98 P(pos)=1.00 Local NLI stance (deberta-v3-large): positive — NLI p=0.80 GPT-6.1 Sol (LIVE): positive — Calls a scene hilarious, expressing enjoyment of the film's action. Claude Opus 5.5: positive — Commenter enjoyed a specific scene, calling it hilarious, which signals they had fun with the film. Claude Sonnet 5.5: positive — Commenter found a specific scene hilarious, implying enjoyment of the film, though no explicit overall recommendation. Gemini 4 Argon: positive — The commenter expresses strong amusement at a specific scene in the film. Grok 4.7: positive — Calls a specific scene hilarious, showing enjoyment of the film. GPT-6 Luna (cheap): no_verdict — Finds a scene hilarious but gives no judgment of the film overall. Claude Haiku 5.5 (cheap): positive — The commenter finds a scene 'fucking hilarious', an enthusiastic reaction that leans toward enjoyment of the film. Gemini 3.5 Flash-Lite (cheap): no_verdict — Commenting on a funny specific scene is not a clear verdict on the whole film. | + | + | + | + | + | + | + | + | 0 | + | 0 |
| thread #72 w 0.117 Ending was abit abrupt, but I was very entertained throughout the journey. I loved how he was moving level to level and buffing up each time with new weapons. VADER (old v1): positive — VADER 0.757 siebert+opinion (previous live): positive — P(op)=0.55 P(pos)=1.00 Local NLI stance (deberta-v3-large): no_verdict — NLI p=0.63 GPT-6.1 Sol (LIVE): positive — Despite an abrupt ending, they were very entertained and loved the level progression and weapon upgrades. Claude Opus 5.5: positive — Minor complaint about the abrupt ending, but very entertained overall and loved the level progression. Claude Sonnet 5.5: positive — Minor complaint about abrupt ending, but overall very entertained and loved the level progression; leans clearly positive. Gemini 4 Argon: positive — Despite noting an abrupt ending, the commenter was very entertained throughout and loved the progression. Grok 4.7: positive — Minor ending critique, but overall entertained and loved the progression. GPT-6 Luna (cheap): positive — They found it entertaining and loved the progression, despite a mildly abrupt ending. Claude Haiku 5.5 (cheap): positive — Commenter says they were very entertained and loved the level-by-level progression, despite noting an abrupt ending; overall lean is positive. Gemini 3.5 Flash-Lite (cheap): positive — The commenter stated they were very entertained throughout and loved aspects of the movie despite a minor flaw. | + | + | 0 | + | + | + | + | + | + | + | + |
| thread #73 w 0.116 It is interesting that it had a bleak ending considering that Resident Evil games usually have the protagonist winning and saving the day. Even Villiage had the protagonist save the day and it is the most bittersweet of all Resident Evil endings VADER (old v1): negative — VADER -0.272 siebert+opinion (previous live): positive — P(op)=0.90 P(pos)=0.99 Local NLI stance (deberta-v3-large): mixed — NLI p=0.35 GPT-6.1 Sol (LIVE): no_verdict — Compares the film’s bleak ending with the games without expressing approval or disapproval of the film. Claude Opus 5.5: no_verdict — Observes the bleak ending compared to the games without judging the film's quality. Claude Sonnet 5.5: no_verdict — Comments on the ending's bleakness compared to the games' endings; no judgement of whether the film is good or bad. Gemini 4 Argon: no_verdict — The comment makes a neutral observation comparing the movie's ending to the video games without judging the film's quality. Grok 4.7: no_verdict — Observes the bleak ending versus games; no film judgement. GPT-6 Luna (cheap): no_verdict — Compares the film’s bleak ending with the games’ endings without judging the film itself. Claude Haiku 5.5 (cheap): no_verdict — Compares the film's bleak ending to game conventions and Village's ending; no clear judgement of the film itself. Gemini 3.5 Flash-Lite (cheap): no_verdict — Discusses plot and compares endings to the video games without judging the film itself. | − | + | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| thread #74 w 0.115 I kind of got the vibe/message 'yeaaaah this guy ain't MC material'. More if the player was in the game rather than the actual mc's VADER (old v1): neutral — VADER 0.0 siebert+opinion (previous live): negative — P(op)=0.72 P(pos)=0.00 Local NLI stance (deberta-v3-large): mixed — NLI p=0.44 GPT-6.1 Sol (LIVE): no_verdict — Interprets a character's role and the film's message without judging the film's quality. Claude Opus 5.5: no_verdict — Comments on a character not seeming like main-character material; no judgement of the film's quality. Claude Sonnet 5.5: no_verdict — Comment discusses a character's vibe or protagonist suitability without clearly judging whether the film is good or recommending it. Gemini 4 Argon: no_verdict — The comment merely discusses a character's role and vibe without expressing a positive or negative opinion on the movie. Grok 4.7: no_verdict — Character discussion only; no judgement of the film. GPT-6 Luna (cheap): no_verdict — Discusses the protagonist's perceived role and game-player perspective without judging the film. Claude Haiku 5.5 (cheap): no_verdict — VADER 0.0 Gemini 3.5 Flash-Lite (cheap): no_verdict — The comment discusses characters and game vibe without expressing a clear verdict on the film. | 0 | − | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| thread #75 w 0.115 Yes, that was my feeling too. Like, the movie's main character isn't the game's main character. The game starts after he has already failed, he's just one of the voice logs that you find along the way; the backstory to one of the monsters. VADER (old v1): positive — VADER 0.34 siebert+opinion (previous live): negative — P(op)=0.56 P(pos)=0.00 Local NLI stance (deberta-v3-large): no_verdict — NLI p=0.49 GPT-6.1 Sol (LIVE): no_verdict — Discusses differences between the movie’s protagonist and the game’s backstory without judging the film. Claude Opus 5.5: no_verdict — Compares the movie's protagonist to the game's backstory without judging the film's quality. Claude Sonnet 5.5: no_verdict — Discusses the movie's departure from the game's protagonist and backstory; no clear overall judgement of the film's quality. Gemini 4 Argon: no_verdict — The comment only discusses the character's role and compares the plot to the games without judging the film's quality. Grok 4.7: no_verdict — Compares film and game characters; no judgement of the movie. GPT-6 Luna (cheap): no_verdict — Discusses how the movie's main character differs from the game's, without judging the film. Claude Haiku 5.5 (cheap): no_verdict — Discusses plot and game-versus-film character differences without judging whether the film is good or worth watching. Gemini 3.5 Flash-Lite (cheap): no_verdict — The commenter discusses plot details and character origins without expressing a clear judgment on the film. | + | − | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| thread #76 w 0.114 Cregger had a nice cameo as the scientist telling Bryan to take the stairs lol VADER (old v1): positive — VADER 0.681 siebert+opinion (previous live): neutral — P(op)=0.19 P(pos)=1.00 Local NLI stance (deberta-v3-large): no_verdict — NLI p=0.85 GPT-6.1 Sol (LIVE): no_verdict — Comments on a nice cameo without expressing an overall judgment of the film. Claude Opus 5.5: no_verdict — Comments on a cameo appearance without judging the film's overall quality. Claude Sonnet 5.5: no_verdict — Mentions a cameo with amusement but gives no overall judgment of the film or recommendation. Gemini 4 Argon: no_verdict — The comment merely points out a fun director cameo without giving an overall verdict on the film itself. Grok 4.7: no_verdict — Cameo trivia joke, no judgement of the film. GPT-6 Luna (cheap): no_verdict — Mentions a cameo and makes a joke, but gives no opinion on the film. Claude Haiku 5.5 (cheap): no_verdict — Mentions a cameo and a trivia-like plot detail with no judgement of the film's quality. Gemini 3.5 Flash-Lite (cheap): no_verdict — The comment only mentions a cameo without giving any judgment on the film itself. | + | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| thread #77 w 0.113 I was bummed he wasn’t one of the lab guys! VADER (old v1): neutral — VADER 0.0 siebert+opinion (previous live): negative — P(op)=0.96 P(pos)=0.00 Local NLI stance (deberta-v3-large): mixed — NLI p=0.35 GPT-6.1 Sol (LIVE): no_verdict — Expresses disappointment about a character’s role, not an overall judgment of the film. Claude Opus 5.5: no_verdict — Minor disappointment about casting/character detail, not a judgement of the film overall. Claude Sonnet 5.5: no_verdict — Expresses mild disappointment about a specific actor/character not being in a role, but gives no overall judgement of the film. Gemini 4 Argon: no_verdict — Expresses minor disappointment about a specific casting or character detail without giving an overall verdict on the film. Grok 4.7: no_verdict — Wishes a character were a lab guy; no film judgement. GPT-6 Luna (cheap): no_verdict — Expresses disappointment about a character’s role, not an opinion of the film. Claude Haiku 5.5 (cheap): no_verdict — Comment expresses disappointment about a plot detail (character's role) but gives no clear judgement of the film overall. Gemini 3.5 Flash-Lite (cheap): no_verdict — Comment discusses character expectation, not an overall judgment of the film. | 0 | − | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| thread #78 w 0.113 In my head they killed 2 birds with 1 stone with that ending: If a sequel is green lit then the cure is sitting right there as the main plot point/objective…if there is no sequel, then they gave us an insane and abnormal ending that most people wouldn’t see coming VADER (old v1): negative — VADER -0.599 siebert+opinion (previous live): positive — P(op)=0.94 P(pos)=0.98 Local NLI stance (deberta-v3-large): mixed — NLI p=0.48 GPT-6.1 Sol (LIVE): no_verdict — Discusses how the ending accommodates a sequel or standalone conclusion without clearly judging the film. Claude Opus 5.5: no_verdict — Analyzes the ending's sequel strategy without clearly judging the film's quality; 'insane' ending is descriptive, not a clear verdict. Claude Sonnet 5.5: positive — Praises the ending as clever and surprising ('insane and abnormal', 'killed 2 birds with 1 stone'), implying approval of the film's choices. Gemini 4 Argon: positive — The commenter praises the cleverness of the ending as 'insane and abnormal' and effective. Grok 4.7: no_verdict — Discusses ending's sequel setup and surprise, no film judgement. GPT-6 Luna (cheap): positive — They praise the ending as a clever, unexpected choice that works whether or not a sequel is made. Claude Haiku 5.5 (cheap): no_verdict — Speculates on the ending's plot setup and sequel possibilities without judging whether the film itself is good or worth watching. Gemini 3.5 Flash-Lite (cheap): no_verdict — The commenter is discussing potential plot outcomes and ending scenarios without expressing a personal judgment on the film. | − | + | 0 | 0 | 0 | + | + | 0 | + | 0 | 0 |
| thread #79 w 0.112 I love how much of the videogamey vibe this movie has, without it ever feeling forced. VADER (old v1): positive — VADER 0.209 siebert+opinion (previous live): positive — P(op)=0.74 P(pos)=1.00 Local NLI stance (deberta-v3-large): no_verdict — NLI p=0.54 GPT-6.1 Sol (LIVE): positive — Praises the movie’s video game vibe and how naturally it is incorporated. Claude Opus 5.5: positive — Commenter loves the film's video-game vibe and says it never feels forced; clear praise. Claude Sonnet 5.5: positive — Commenter explicitly loves the videogame-like vibe of the film and says it never feels forced, which is clear praise. Gemini 4 Argon: positive — The commenter praises the movie's video game vibe and natural execution. Grok 4.7: positive — Praises the film's authentic, unforced videogame vibe. GPT-6 Luna (cheap): positive — They praise the movie’s videogame-like vibe and say it feels natural. Claude Haiku 5.5 (cheap): positive — Commenter praises the film's video-game vibe and says it never feels forced, a clear approval. Gemini 3.5 Flash-Lite (cheap): positive — The commenter praises the movie's vibe and execution. | + | + | 0 | + | + | + | + | + | + | + | + |
| thread #80 w 0.111 The POV shots were so well done VADER (old v1): positive — VADER 0.41 siebert+opinion (previous live): positive — P(op)=0.63 P(pos)=1.00 Local NLI stance (deberta-v3-large): mixed — NLI p=0.35 GPT-6.1 Sol (LIVE): positive — Praises the film’s execution of POV shots. Claude Opus 5.5: positive — Praises the film's POV shots as well done, a positive judgement of its craft. Claude Sonnet 5.5: positive — Praises the POV shots as well done, a favorable judgment of the film's craft, with no criticism offered. Gemini 4 Argon: positive — The commenter directly praises the execution of the film's cinematography (the POV shots). Grok 4.7: positive — Praises the film's well-executed POV shots. GPT-6 Luna (cheap): positive — Praises the film’s POV shots as well done. Claude Haiku 5.5 (cheap): positive — Praises the POV shots as 'so well done', a clear compliment on the film's craft. Gemini 3.5 Flash-Lite (cheap): positive — Praises a specific filmmaking aspect of the movie. | + | + | 0 | + | + | + | + | + | + | + | + |
| thread #81 w 0.11 I think the moral of this movie is that people don’t respect delivery drivers enough. VADER (old v1): positive — VADER 0.477 siebert+opinion (previous live): negative — P(op)=1.00 P(pos)=0.00 Local NLI stance (deberta-v3-large): mixed — NLI p=0.42 GPT-6.1 Sol (LIVE): no_verdict — A joke about the movie’s moral, without expressing approval or criticism of the film. Claude Opus 5.5: no_verdict — Joking interpretation of the movie's moral; no judgement of the film's quality. Claude Sonnet 5.5: no_verdict — A joking remark about the film's supposed moral; no judgment of whether the film is good or bad. Gemini 4 Argon: no_verdict — The comment makes a joke about the plot's moral without expressing an opinion on the film's quality. Grok 4.7: no_verdict — Joke about the movie's moral, no judgement of quality. GPT-6 Luna (cheap): no_verdict — A humorous interpretation of the movie’s message, with no clear opinion about the film itself. Claude Haiku 5.5 (cheap): no_verdict — A joking remark about the movie's moral; it does not express a clear judgement of the film itself. Gemini 3.5 Flash-Lite (cheap): no_verdict — The comment is a joke about the plot and does not express a verdict on the film. | + | − | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| thread #82 w 0.11 I don't what it was, but the final monster transformation in the end was eerily and disturbingly captivating. Top tier monster design. VADER (old v1): negative — VADER -0.502 siebert+opinion (previous live): positive — P(op)=0.88 P(pos)=1.00 Local NLI stance (deberta-v3-large): positive — NLI p=0.90 GPT-6.1 Sol (LIVE): positive — Praises the final monster transformation as captivating and its design as top tier. Claude Opus 5.5: positive — Praises the disturbing final monster transformation as captivating and calls the monster design top tier. Claude Sonnet 5.5: positive — Praises the final monster transformation as captivating and top-tier monster design; disturbing content is framed as admiration for the film. Gemini 4 Argon: positive — The commenter praises the final monster transformation as captivating and calls the monster design top tier. Grok 4.7: positive — Praises the final monster transformation and design as captivating and top tier. GPT-6 Luna (cheap): positive — Calls the final transformation captivating and the monster design top tier, expressing clear praise. Claude Haiku 5.5 (cheap): positive — Praises the final monster design as captivating and top tier, a clear positive judgement of that element of the film. Gemini 3.5 Flash-Lite (cheap): positive — Praises the monster design as top tier and captivating. | − | + | + | + | + | + | + | + | + | + | + |
| thread #83 w 0.109 Because he seemed somewhat aware the whole time. All the people seemed somewhat aware as they turned. VADER (old v1): neutral — VADER 0.0 siebert+opinion (previous live): neutral — P(op)=0.06 P(pos)=1.00 Local NLI stance (deberta-v3-large): no_verdict — NLI p=0.58 GPT-6.1 Sol (LIVE): no_verdict — Discusses characters' awareness while turning, without judging the film. Claude Opus 5.5: no_verdict — Discusses plot details about characters' awareness while turning; no judgement of the film. Claude Sonnet 5.5: no_verdict — Comment discusses plot/character details about people turning, with no judgement of the film's quality or recommendation. Gemini 4 Argon: no_verdict — Purely plot discussion about the characters turning, with no expressed opinion on the film's quality. Grok 4.7: no_verdict — Plot observation about character awareness, no film judgement. GPT-6 Luna (cheap): no_verdict — Observes the characters’ awareness while turning, without expressing an opinion about the film. Claude Haiku 5.5 (cheap): no_verdict — Discusses a plot detail about characters being aware while turning; no judgement of the film's quality. Gemini 3.5 Flash-Lite (cheap): no_verdict — The comment discusses a plot detail about characters turning without expressing any opinion on the film itself. | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| thread #84 w 0.108 That was my favorite. Him being aware of what's happening and not being able to stop it. Terrifying. VADER (old v1): positive — VADER 0.382 siebert+opinion (previous live): positive — P(op)=0.94 P(pos)=1.00 Local NLI stance (deberta-v3-large): positive — NLI p=0.80 GPT-6.1 Sol (LIVE): positive — Calls the scene their favorite and praises its terrifying helplessness. Claude Opus 5.5: positive — Calls a scene their favorite and praises it as terrifying, which is admiration of the film. Claude Sonnet 5.5: positive — Calls something their favorite and praises it as terrifying; praise of scary content counts as a positive verdict on the film. Gemini 4 Argon: positive — The commenter calls the scene/film their favorite and praises it as effectively terrifying. Grok 4.7: positive — Calls a terrifying scene their favorite, praising the horror. GPT-6 Luna (cheap): positive — Calls it their favorite and describes the scene as terrifying in an appreciative way. Claude Haiku 5.5 (cheap): positive — Commenter calls it their favorite and praises the terrifying scene, a clear endorsement of the film. Gemini 3.5 Flash-Lite (cheap): positive — The commenter calls the movie their favorite and praises its terrifying elements. | + | + | + | + | + | + | + | + | + | + | + |
| thread #85 w 0.108 The shotgun shells being in a game-accurate green box made me very happy. VADER (old v1): positive — VADER 0.612 siebert+opinion (previous live): positive — P(op)=0.43 P(pos)=1.00 Local NLI stance (deberta-v3-large): no_verdict — NLI p=0.47 GPT-6.1 Sol (LIVE): positive — Expresses delight at the film’s game-accurate detail, offering clear praise. Claude Opus 5.5: no_verdict — Enjoys a game-accurate prop detail, but gives no overall judgement of the film's quality. Claude Sonnet 5.5: no_verdict — Expresses delight at one small game-accurate detail, but gives no overall verdict on the film or recommendation. Gemini 4 Argon: positive — The commenter expresses happiness over a game-accurate detail in the film, with no criticism. Grok 4.7: no_verdict — Happy about one game-accurate prop detail, not a film judgement. GPT-6 Luna (cheap): positive — They enjoyed the game-accurate detail in the movie. Claude Haiku 5.5 (cheap): no_verdict — Comment notes a game-accurate detail pleasing them, but gives no clear judgement of the film overall. Gemini 3.5 Flash-Lite (cheap): positive — The commenter expresses happiness regarding a detail in the film. | + | + | 0 | + | 0 | 0 | + | 0 | + | 0 | + |
| thread #86 w 0.107 i also saw 2 first aid sprays in the barn before he dropped down! VADER (old v1): neutral — VADER 0.0 siebert+opinion (previous live): negative — P(op)=0.47 P(pos)=0.01 Local NLI stance (deberta-v3-large): no_verdict — NLI p=0.49 GPT-6.1 Sol (LIVE): no_verdict — The commenter points out an observed detail without judging the film. Claude Opus 5.5: no_verdict — Trivia/easter egg observation about first aid sprays; no judgement of the film's quality. Claude Sonnet 5.5: no_verdict — Commenter points out an in-game detail (first aid sprays in the barn); no judgement of the film's quality. Gemini 4 Argon: no_verdict — The comment simply points out an Easter egg in a scene without expressing an opinion on the film. Grok 4.7: no_verdict — Scene detail observation with no film judgement. GPT-6 Luna (cheap): no_verdict — Points out an object spotted in a scene without expressing an opinion about the film. Claude Haiku 5.5 (cheap): no_verdict — Just notes a detail about items in a scene; no judgement of the film's quality. Gemini 3.5 Flash-Lite (cheap): no_verdict — The commenter points out a detail in a scene without expressing any judgment or opinion about the film. | 0 | − | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| thread #87 w 0.107 I got the bad ending in the movie. Does someone know how to get the good ending? I must have missed a key VADER (old v1): negative — VADER -0.421 siebert+opinion (previous live): negative — P(op)=0.77 P(pos)=0.00 Local NLI stance (deberta-v3-large): negative — NLI p=0.51 GPT-6.1 Sol (LIVE): no_verdict — Jokes about video game endings and missing a key, without judging the film. Claude Opus 5.5: no_verdict — Joke treating the movie like the video game; no judgement of the film's quality. Claude Sonnet 5.5: no_verdict — Joking question about endings, as if the film were a game; no judgement of the film's quality or recommendation. Gemini 4 Argon: no_verdict — The comment is a joke comparing the movie to a video game without expressing an opinion on the film's quality. Grok 4.7: no_verdict — Asks about endings and a missed key; no film judgement. GPT-6 Luna (cheap): no_verdict — A joke about the movie having alternate endings, with no clear opinion of the film itself. Claude Haiku 5.5 (cheap): no_verdict — Asks about endings in a game-style joke/confusion; no judgement of the film's quality or whether they'd recommend it. Gemini 3.5 Flash-Lite (cheap): no_verdict — The commenter is making a joke about a video game trope applied to a movie. | − | − | − | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| thread #88 w 0.106 Bad ending branch goes all the way back to the tunnel. Sadly, you got to shoot the kid right away. VADER (old v1): negative — VADER -0.827 siebert+opinion (previous live): negative — P(op)=0.98 P(pos)=0.00 Local NLI stance (deberta-v3-large): negative — NLI p=0.96 GPT-6.1 Sol (LIVE): no_verdict — Discusses an ending branch and shooting a character without judging the film. Claude Opus 5.5: no_verdict — Discusses plot/ending branches and a scene without judging the film's quality. Claude Sonnet 5.5: no_verdict — Comment discusses a plot or ending branch, apparently game-like, without judging the film itself as good or bad. Gemini 4 Argon: no_verdict — The comment makes a joke framing the plot as a video game with branching endings, offering no film evaluation. Grok 4.7: no_verdict — Discusses a game ending branch, not a film judgement. GPT-6 Luna (cheap): no_verdict — Describes an ending branch and an in-story action without judging the film. Claude Haiku 5.5 (cheap): no_verdict — Discusses a game's bad ending branch and plot choices without judging the film itself. Gemini 3.5 Flash-Lite (cheap): no_verdict — Discusses a video game branch or plot mechanic rather than giving a verdict on the film. | − | − | − | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| thread #89 w 0.105 “It was you wasn’t it?” Had me choking on my popcorn VADER (old v1): negative — VADER -0.459 siebert+opinion (previous live): negative — P(op)=0.94 P(pos)=0.00 Local NLI stance (deberta-v3-large): positive — NLI p=0.39 GPT-6.1 Sol (LIVE): positive — The quoted moment amused them enough to choke on their popcorn, conveying enjoyment. Claude Opus 5.5: no_verdict — Quotes a line and reacts to it, possibly with laughter, but gives no clear judgement of the film. Claude Sonnet 5.5: no_verdict — Quotes a line and says it made them choke on popcorn (likely laughter or shock), but gives no explicit judgement of the film. Gemini 4 Argon: no_verdict — The comment merely reacts to a single quote without offering an overall judgment on the film. Grok 4.7: no_verdict — Quote reaction only; no overall film judgement. GPT-6 Luna (cheap): positive — The commenter says the moment made them choke on their popcorn, indicating they enjoyed it. Claude Haiku 5.5 (cheap): no_verdict — Quotes a line and jokes about it; no judgement of the film's quality or recommendation. Gemini 3.5 Flash-Lite (cheap): no_verdict — Quoting a funny or memorable line from the movie without judging the film itself. | − | − | + | + | 0 | 0 | 0 | 0 | + | 0 | 0 |
| thread #90 w 0.105 Some of the jump scares in this movie were so effective I felt like that kid play RE3 in 1999 again, 9/10. VADER (old v1): positive — VADER 0.757 siebert+opinion (previous live): positive — P(op)=1.00 P(pos)=1.00 Local NLI stance (deberta-v3-large): positive — NLI p=0.71 GPT-6.1 Sol (LIVE): positive — Praises the effective jump scares and gives the film a 9/10 rating. Claude Opus 5.5: positive — Praises effective jump scares and nostalgic feel, rates the film 9/10. Claude Sonnet 5.5: positive — Praises effective jump scares with nostalgic enthusiasm and gives the film a 9/10 rating. Gemini 4 Argon: positive — The commenter praises the effective jump scares and explicitly gives the movie a 9/10 rating. Grok 4.7: positive — Praises effective jump scares and rates the film 9/10. GPT-6 Luna (cheap): positive — They praise the jump scares as effective and rate the movie 9/10. Claude Haiku 5.5 (cheap): positive — Commenter praises the jump scares as effective and gives the film a 9/10. Gemini 3.5 Flash-Lite (cheap): positive — High rating of 9/10 and praise for effective jump scares. | + | + | + | + | + | + | + | + | + | + | + |
| thread #91 w 0.104 Does this also count as a Katamari movie? VADER (old v1): neutral — VADER 0.0 siebert+opinion (previous live): neutral — P(op)=0.11 P(pos)=0.12 Local NLI stance (deberta-v3-large): no_verdict — NLI p=0.39 GPT-6.1 Sol (LIVE): no_verdict — A joke comparing the movie to Katamari, without expressing an opinion on its quality. Claude Opus 5.5: no_verdict — Joke question about the film's genre; no judgement of the film's quality. Claude Sonnet 5.5: no_verdict — A joking question about whether the film counts as a Katamari movie; it gives no judgement of the film. Gemini 4 Argon: no_verdict — The comment is a joke comparing the movie to a video game without expressing an opinion on its quality. Grok 4.7: no_verdict — Joke reference only; no judgment of the film. GPT-6 Luna (cheap): no_verdict — A joke about the movie, with no expressed opinion on its quality. Claude Haiku 5.5 (cheap): no_verdict — A question about whether the film counts as a Katamari movie; no judgement of the film's quality expressed. Gemini 3.5 Flash-Lite (cheap): no_verdict — The comment is a joke/question about the plot and unrelated media, offering no judgment on the film. | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| thread #92 w 0.104 “You never shoot a gun in a car, you could hit the engine!” Got one of the biggest laughs I’ve heard in a theater in a long time VADER (old v1): positive — VADER 0.762 siebert+opinion (previous live): neutral — P(op)=0.16 P(pos)=1.00 Local NLI stance (deberta-v3-large): no_verdict — NLI p=0.50 GPT-6.1 Sol (LIVE): positive — Highlights a line earning an exceptionally big theater laugh, praising the film’s humor. Claude Opus 5.5: no_verdict — Quotes a line and reports audience laughter; unclear whether praise or mockery, no judgement of the film. Claude Sonnet 5.5: no_verdict — Quotes a line and notes the audience laughed, but gives no clear judgement of the film itself; the laugh could be at or with it. Gemini 4 Argon: no_verdict — Quotes a line and describes the theater audience's reaction without giving a personal verdict on the film. Grok 4.7: positive — Highlights a line that got huge theater laughs, signaling enjoyment. GPT-6 Luna (cheap): positive — They praise the scene’s humor, noting it got one of the biggest laughs they’ve heard in a theater. Claude Haiku 5.5 (cheap): positive — Commenter praises the film by saying the line got one of the biggest laughs they've heard in a theater in a long time. Gemini 3.5 Flash-Lite (cheap): positive — Praising a humorous moment and audience reaction indicates an enjoyable experience. | + | 0 | 0 | + | 0 | 0 | 0 | + | + | + | + |
| thread #93 w 0.103 My man was so panicked he accidentally picked the Drop option on his shotgun shells. VADER (old v1): negative — VADER -0.824 siebert+opinion (previous live): neutral — P(op)=0.11 P(pos)=0.00 Local NLI stance (deberta-v3-large): no_verdict — NLI p=0.40 GPT-6.1 Sol (LIVE): no_verdict — Jokes about a character's panicked actions without judging the film. Claude Opus 5.5: no_verdict — Joke about a character's panicked action; no judgement of the film's quality. Claude Sonnet 5.5: no_verdict — A joke about a character's panicked mistake in a scene, with no judgement of the film itself. Gemini 4 Argon: no_verdict — The comment makes a video game joke about a scene without expressing an opinion on the movie. Grok 4.7: no_verdict — Joke about a panicked character; no judgement of the film. GPT-6 Luna (cheap): no_verdict — Jokes about a character's actions without expressing an opinion about the film. Claude Haiku 5.5 (cheap): no_verdict — Joke about a character's panicked mishap; no judgement of the film itself expressed. Gemini 3.5 Flash-Lite (cheap): no_verdict — The comment is a joke about a scene without judging the film itself. | − | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| thread #94 w 0.103 I like to imagine Leon or Claire picking up the antidote on the ground somewhere, and find his audio log. RIP random NPC zombie. VADER (old v1): positive — VADER 0.361 siebert+opinion (previous live): neutral — P(op)=0.13 P(pos)=1.00 Local NLI stance (deberta-v3-large): no_verdict — NLI p=0.62 GPT-6.1 Sol (LIVE): no_verdict — Imagines a character scenario and jokes about an NPC without judging the film. Claude Opus 5.5: no_verdict — Joke imagining game characters finding the antidote; no judgement of the film. Claude Sonnet 5.5: no_verdict — Playful fan speculation about game characters and a zombie; no judgement of the film itself. Gemini 4 Argon: no_verdict — The comment makes a joke/reference connecting the plot to the video games without judging the film. Grok 4.7: no_verdict — Fan speculation and joke; no judgement of the film. GPT-6 Luna (cheap): no_verdict — Imagines a character finding an antidote and audio log; offers no opinion on the film. Claude Haiku 5.5 (cheap): no_verdict — Playful fan theory about a game-related scenario; no judgement of the film itself expressed. Gemini 3.5 Flash-Lite (cheap): no_verdict — The commenter makes a joke/theory about the video game characters and plot, not judging the movie. | + | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| thread #95 w 0.102 LOVE ❤️ Downer ending but this is very much the side story of a character you read about in notes/recordings before you face them as a mini boss VADER (old v1): positive — VADER 0.453 siebert+opinion (previous live): neutral — P(op)=0.13 P(pos)=1.00 Local NLI stance (deberta-v3-large): mixed — NLI p=0.85 GPT-6.1 Sol (LIVE): positive — Explicitly says they love it, despite the downer ending. Claude Opus 5.5: positive — Opens with 'LOVE'; the downer ending is described, not criticised, and framed as fitting lore. Claude Sonnet 5.5: positive — Opens with 'LOVE' and the 'downer ending' remark is descriptive, not a criticism; overall enthusiastic about the film. Gemini 4 Argon: positive — The commenter explicitly states 'LOVE ❤️' and appreciates the film's narrative structure despite a downer ending. Grok 4.7: positive — Explicit love for the film despite a downer ending. GPT-6 Luna (cheap): positive — The commenter explicitly loves it; the downer ending is descriptive, not criticism. Claude Haiku 5.5 (cheap): positive — Commenter says 'LOVE ❤️' despite noting a downer ending; clear enthusiasm for the film overall. Gemini 3.5 Flash-Lite (cheap): positive — The commenter expresses clear enjoyment with 'LOVE ❤️'. | + | 0 | 0 | + | + | + | + | + | + | + | + |
| thread #96 w 0.102 Zach Gregger continues his winning streak! I love the title drop 20mins into the movie Austin Abrams is perfect as the courier. He behaves how a normal person would if they got pulled into this chaos. His crash outs and disbelief provided the perfect balance of humour The… VADER (old v1): positive — VADER 0.973 siebert+opinion (previous live): positive — P(op)=1.00 P(pos)=1.00 Local NLI stance (deberta-v3-large): positive — NLI p=0.97 GPT-6.1 Sol (LIVE): positive — Enthusiastically praises performances, cinematography, action, and humor, calling it a fun, gory crowd pleaser despite one criticism. Claude Opus 5.5: positive — Extensive praise for acting, cinematography, and set pieces; calls it a fun, gory crowd pleaser despite one minor gripe. Claude Sonnet 5.5: positive — Enthusiastic praise of acting, cinematography, and set pieces, concluding it's a fun, gory crowd pleaser; only a minor quibble about Bryan's fate. Gemini 4 Argon: positive — Praises the direction, acting, cinematography, lighting, and third act, calling it a "very fun and gory crowd pleaser." Grok 4.7: positive — Praises direction, cast, visuals, and gore; calls it a fun crowd pleaser despite one plot gripe. GPT-6 Luna (cheap): positive — They praise the cinematography, performances, set pieces, and call it a fun, gory crowd-pleaser despite one story choice. Claude Haiku 5.5 (cheap): positive — Enthusiastic praise throughout (cinematography, acting, action, ending), closing with 'very fun... crowd pleaser'; minor gripe doesn't change overall lean. Gemini 3.5 Flash-Lite (cheap): positive — Praises the director, cinematography, action, and calls it a very fun and gory crowd pleaser. | + | + | + | + | + | + | + | + | + | + | + |
| thread #97 w 0.101 “Aw man, not my fucking jays!” This was a ton of fun in the theaters with a packed crowd. VADER (old v1): positive — VADER 0.556 siebert+opinion (previous live): positive — P(op)=0.96 P(pos)=1.00 Local NLI stance (deberta-v3-large): mixed — NLI p=0.50 GPT-6.1 Sol (LIVE): positive — Calls the movie a ton of fun, especially with a packed theater crowd. Claude Opus 5.5: positive — Quotes a line and says the film was a ton of fun with a packed theater crowd. Claude Sonnet 5.5: positive — Commenter says it was a ton of fun in theaters with a packed crowd, indicating a clear enjoyment of the film. Gemini 4 Argon: positive — The commenter explicitly states the movie was "a ton of fun" to watch in theaters. Grok 4.7: positive — Explicitly calls the theatrical experience a ton of fun. GPT-6 Luna (cheap): positive — Calls it a ton of fun, indicating they enjoyed the film. Claude Haiku 5.5 (cheap): positive — Commenter says it was 'a ton of fun' in theaters with a packed crowd, clearly enjoying the film. Gemini 3.5 Flash-Lite (cheap): positive — The commenter states the movie was 'a ton of fun in the theaters'. | + | + | 0 | + | + | + | + | + | + | + | + |
| thread #98 w 0.101 Very on brand for Umbrella to allow an infected into the lab and to also not have multiple versions of the base for a cure on hand, shocked Bryan didn’t need to upgrade a keycard as he progressed up the hospital VADER (old v1): negative — VADER -0.103 siebert+opinion (previous live): positive — P(op)=0.82 P(pos)=0.99 Local NLI stance (deberta-v3-large): mixed — NLI p=0.53 GPT-6.1 Sol (LIVE): no_verdict — Jokes about Umbrella’s incompetence and game-like keycard progression without judging the film. Claude Opus 5.5: no_verdict — Joking about plot logic and game references; no clear judgement of the film's quality. Claude Sonnet 5.5: no_verdict — Joking in-universe commentary about Umbrella's incompetence and game-like mechanics; no explicit judgement of the film's quality. Gemini 4 Argon: no_verdict — The comment makes a joke about Resident Evil franchise tropes without expressing an opinion on the movie's quality. Grok 4.7: negative — Sarcastically mocks the film’s illogical plot and video-game tropes. GPT-6 Luna (cheap): negative — They criticize the film’s implausible plot logic and progression, implying dissatisfaction with its writing. Claude Haiku 5.5 (cheap): no_verdict — Plot-based joking about Umbrella's and characters' decisions; no explicit judgement of the film's quality or recommendation. Gemini 3.5 Flash-Lite (cheap): no_verdict — The comment discusses plot logic and makes a joke about video game tropes without judging the film itself. | − | + | 0 | 0 | 0 | 0 | 0 | − | − | 0 | 0 |