<a href=\"https://arxiv.org/pdf/2608.23172\" rel=\"nofollow\">https://arxiv.org/pdf/2608.23172</a></p>\n<p>Large-scale vision-language models (VLMs) have demonstrated remarkable versatility across a wide range of multimodal tasks. However, understanding humor remains challenging because humorous content often depends on subtle interactions among entities, events, context, and implicit relationships across image and text modalities. These interactions can involve complex chains of reasoning that are difficult to capture through conventional prompting or linear chain-of-thought reasoning. In this work, we propose CaRGo-T (Causal Reasoning Graph-of-Thought), a reasoning framework that represents the causal and contextual relationships underlying multimodal humor as a lightweight graph-based reasoning structure. The graph is serialized into a code-based representation generated by a VLM, which can subsequently be interpreted by the same or a different VLM to produce the final prediction in zero-shot or in-context learning settings. We evaluate CaRGo-T on humor understanding and humor detection across four datasets spanning diverse forms of comedic content, including satire, sarcasm, and memes. Experiments with state-of-the-art commercial and open-source VLMs show that CaRGo-T consistently improves performance over existing reasoning-based baselines, achieving gains of approximately 1-20% on humor understanding and 1-3% on humor detection. Further analysis using mutual information indicates that the reasoning representations produced by CaRGo-T contain more information relevant to the target output than those generated by baseline reasoning approaches. Code is available at <a href=\"https://github.com/abhi1nandy2/CaRGo-T\" rel=\"nofollow\">https://github.com/abhi1nandy2/CaRGo-T</a>.</p>\n","updatedAt":"2026-08-28T02:45:02.738Z","author":{"_id":"5f89da6c5d083370c711f37c","avatarUrl":"/avatars/c2f17a4a636973817fd5da2ae6dbaac3.svg","fullname":"Abhilash Nandy","name":"abhi1nandy2","type":"user","isPro":false,"isHf":false,"isHfAdmin":false,"isMod":false,"followerCount":9,"isUserFollowing":false}},"numEdits":0,"identifiedLanguage":{"language":"en","probability":0.9224433898925781},"editors":["abhi1nandy2"],"editorAvatarUrls":["/avatars/c2f17a4a636973817fd5da2ae6dbaac3.svg"],"reactions":[],"isReport":false}}],"primaryEmailConfirmed":false,"paper":{"id":"2608.23172","authors":[{"_id":"6a90f4d0a64059bab69c3616","name":"Abhilash Nandy","hidden":false},{"_id":"6a90f4d0a64059bab69c3617","name":"Rahul Seetharaman","hidden":false},{"_id":"6a90f4d0a64059bab69c3618","name":"Aman Bansal","hidden":false},{"_id":"6a90f4d0a64059bab69c3619","name":"Rounak Saha","hidden":false},{"_id":"6a90f4d0a64059bab69c361a","name":"Manav Nitin Kapadnis","hidden":false},{"_id":"6a90f4d0a64059bab69c361b","name":"Millon Madhur Das","hidden":false},{"_id":"6a90f4d0a64059bab69c361c","name":"Pawan Goyal","hidden":false},{"_id":"6a90f4d0a64059bab69c361d","name":"Niloy Ganguly","hidden":false}],"mediaUrls":["https://cdn-uploads.huggingface.co/production/uploads/5f89da6c5d083370c711f37c/BhNseCMu0qK1vhqRowLp5.jpeg","https://cdn-uploads.huggingface.co/production/uploads/5f89da6c5d083370c711f37c/YHnUj9VrJUpN9USGdeFkS.jpeg"],"publishedAt":"2026-08-24T00:00:00.000Z","submittedOnDailyAt":"2026-08-28T00:00:00.000Z","title":"CaRGo-T: Causal Reasoning Graph-of-Thought improves Multimodal Humor Comprehension","submittedOnDailyBy":{"_id":"5f89da6c5d083370c711f37c","avatarUrl":"/avatars/c2f17a4a636973817fd5da2ae6dbaac3.svg","isPro":false,"fullname":"Abhilash Nandy","user":"abhi1nandy2","type":"user","name":"abhi1nandy2"},"summary":"Large-scale vision-language models (VLMs) have demonstrated remarkable versatility across a wide range of multimodal tasks. However, understanding humor remains challenging because humorous content often depends on subtle interactions among entities, events, context, and implicit relationships across image and text modalities. These interactions can involve complex chains of reasoning that are difficult to capture through conventional prompting or linear chain-of-thought reasoning. In this work, we propose CaRGo-T (Causal Reasoning Graph-of-Thought), a reasoning framework that represents the causal and contextual relationships underlying multimodal humor as a lightweight graph-based reasoning structure. The graph is serialized into a code-based representation generated by a VLM, which can subsequently be interpreted by the same or a different VLM to produce the final prediction in zero-shot or in-context learning settings. We evaluate CaRGo-T on humor understanding and humor detection across four datasets spanning diverse forms of comedic content, including satire, sarcasm, and memes. Experiments with state-of-the-art commercial and open-source VLMs show that CaRGo-T consistently improves performance over existing reasoning-based baselines, achieving gains of approximately 1-20% on humor understanding and 1-3% on humor detection. Further analysis using mutual information indicates that the reasoning representations produced by CaRGo-T contain more information relevant to the target output than those generated by baseline reasoning approaches. Code is available at https://github.com/abhi1nandy2/CaRGo-T.","upvotes":2,"discussionId":"6a90f4d0a64059bab69c361e","githubRepo":"https://github.com/abhi1nandy2/CaRGo-T","githubRepoAddedBy":"user","ai_summary":"CaRGo-T improves multimodal humor understanding by modeling causal relationships as graph-based reasoning structures interpreted by vision-language models.","ai_keywords":["vision-language models","causal reasoning","graph-of-thought","multimodal humor","zero-shot","in-context learning","mutual information"],"ai_summary_model":"thinkingmachines/Inkling-Small","githubStars":0},"canReadDatabase":false,"canManagePapers":false,"canSubmit":false,"hasHfLevelAccess":false,"upvoted":false,"upvoters":[{"_id":"5f89da6c5d083370c711f37c","avatarUrl":"/avatars/c2f17a4a636973817fd5da2ae6dbaac3.svg","isPro":false,"fullname":"Abhilash Nandy","user":"abhi1nandy2","type":"user"},{"_id":"69f1e68fadabd52c938caf54","avatarUrl":"/avatars/2f22c63e3124fa7e443eed89b91a0f35.svg","isPro":false,"fullname":"anon","user":"anony-mouse1","type":"user"}],"acceptLanguages":["en"],"dailyPaperRank":0,"markdownContentUrl":"https://huggingface.co/buckets/huggingchat/papers-content/resolve/2608/2608.23172.md","query":{}}">
CaRGo-T: Causal Reasoning Graph-of-Thought improves Multimodal Humor Comprehension
Abstract
CaRGo-T improves multimodal humor understanding by modeling causal relationships as graph-based reasoning structures interpreted by vision-language models.
Large-scale vision-language models (VLMs) have demonstrated remarkable versatility across a wide range of multimodal tasks. However, understanding humor remains challenging because humorous content often depends on subtle interactions among entities, events, context, and implicit relationships across image and text modalities. These interactions can involve complex chains of reasoning that are difficult to capture through conventional prompting or linear chain-of-thought reasoning. In this work, we propose CaRGo-T (Causal Reasoning Graph-of-Thought), a reasoning framework that represents the causal and contextual relationships underlying multimodal humor as a lightweight graph-based reasoning structure. The graph is serialized into a code-based representation generated by a VLM, which can subsequently be interpreted by the same or a different VLM to produce the final prediction in zero-shot or in-context learning settings. We evaluate CaRGo-T on humor understanding and humor detection across four datasets spanning diverse forms of comedic content, including satire, sarcasm, and memes. Experiments with state-of-the-art commercial and open-source VLMs show that CaRGo-T consistently improves performance over existing reasoning-based baselines, achieving gains of approximately 1-20% on humor understanding and 1-3% on humor detection. Further analysis using mutual information indicates that the reasoning representations produced by CaRGo-T contain more information relevant to the target output than those generated by baseline reasoning approaches. Code is available at https://github.com/abhi1nandy2/CaRGo-T.
Community
https://arxiv.org/pdf/2608.23172
Large-scale vision-language models (VLMs) have demonstrated remarkable versatility across a wide range of multimodal tasks. However, understanding humor remains challenging because humorous content often depends on subtle interactions among entities, events, context, and implicit relationships across image and text modalities. These interactions can involve complex chains of reasoning that are difficult to capture through conventional prompting or linear chain-of-thought reasoning. In this work, we propose CaRGo-T (Causal Reasoning Graph-of-Thought), a reasoning framework that represents the causal and contextual relationships underlying multimodal humor as a lightweight graph-based reasoning structure. The graph is serialized into a code-based representation generated by a VLM, which can subsequently be interpreted by the same or a different VLM to produce the final prediction in zero-shot or in-context learning settings. We evaluate CaRGo-T on humor understanding and humor detection across four datasets spanning diverse forms of comedic content, including satire, sarcasm, and memes. Experiments with state-of-the-art commercial and open-source VLMs show that CaRGo-T consistently improves performance over existing reasoning-based baselines, achieving gains of approximately 1-20% on humor understanding and 1-3% on humor detection. Further analysis using mutual information indicates that the reasoning representations produced by CaRGo-T contain more information relevant to the target output than those generated by baseline reasoning approaches. Code is available at https://github.com/abhi1nandy2/CaRGo-T.
Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.
Tap or paste here to upload images
Cite arxiv.org/abs/2608.23172 in a model README.md to link it from this page.
Cite arxiv.org/abs/2608.23172 in a dataset README.md to link it from this page.
Cite arxiv.org/abs/2608.23172 in a Space README.md to link it from this page.
Discussion (0)
Sign in to join the discussion. Free account, 30 seconds — email code or GitHub.
Sign in →No comments yet. Sign in and be the first to say something.