This survey examines visual humor understanding across memes, cartoons, and comics through a capability-centric hierarchy comprising recognition, interpretation/reasoning, and generation. This taxonomy reorganizes a fragmented literature by focusing on the capabilities that benchmarks and models actually evaluate. Beyond synthesizing prior work, the authors conduct a cross-benchmark evaluation of recent MLLMs, showing that while current models perform reasonably well on visual recognition, they still lag far behind humans on interpretation-intensive tasks. The results also reveal that model rankings vary substantially across humor capabilities and that explicit reasoning variants do not consistently improve performance.</p>\n","updatedAt":"2026-07-22T09:50:15.689Z","author":{"_id":"63999a6fe657365725d0d0a4","avatarUrl":"/avatars/99736de1bc0d5decf4a6eda86e3c7937.svg","fullname":"Derek Zhe Hu","name":"zhehuderek","type":"user","isPro":false,"isHf":false,"isHfAdmin":false,"isMod":false,"followerCount":1,"isUserFollowing":false}},"numEdits":0,"identifiedLanguage":{"language":"en","probability":0.9057374000549316},"editors":["zhehuderek"],"editorAvatarUrls":["/avatars/99736de1bc0d5decf4a6eda86e3c7937.svg"],"reactions":[],"isReport":false}}],"primaryEmailConfirmed":false,"paper":{"id":"2607.19011","authors":[{"_id":"6a60923c1810d97f8fc0ce34","name":"Tuo Liang","hidden":false},{"_id":"6a60923c1810d97f8fc0ce35","name":"Zhe Hu","hidden":false},{"_id":"6a60923c1810d97f8fc0ce36","name":"Disheng Liu","hidden":false},{"_id":"6a60923c1810d97f8fc0ce37","name":"Jing Li","hidden":false},{"_id":"6a60923c1810d97f8fc0ce38","name":"Yu Yin","hidden":false}],"publishedAt":"2026-07-21T00:00:00.000Z","submittedOnDailyAt":"2026-07-22T00:00:00.000Z","title":"Computational Humor with Multimodal LLMs: Methods, Datasets, Evaluation, and Challenges","submittedOnDailyBy":{"_id":"63999a6fe657365725d0d0a4","avatarUrl":"/avatars/99736de1bc0d5decf4a6eda86e3c7937.svg","isPro":false,"fullname":"Derek Zhe Hu","user":"zhehuderek","type":"user","name":"zhehuderek"},"summary":"Multimodal humor in memes, cartoons, and comics remains difficult for AI systems because intended meaning depends on non-literal mechanisms, shared cultural knowledge, and communicative intent rather than literal scene description. This survey focuses on visual humor understanding in single-image and multi-panel artifacts, while treating humor generation as an emerging downstream frontier. We position the literature against prior humor, sarcasm, and general MLLM surveys and organize it using a capability-centric hierarchy spanning recognition, interpretation and reasoning, and generation. Under this lens, we synthesize benchmark design, evaluation protocols, and modeling paradigms, tracing the field's shift from task-specific fusion models to large-model approaches based on multimodal alignment, evidence-grounded reasoning, and controlled generation. We conclude by highlighting the main barriers to progress: shortcut-prone evaluation, limited cultural and narrative coverage, weak evidence grounding, and unresolved safety and ownership concerns.","upvotes":0,"discussionId":"6a60923c1810d97f8fc0ce39"},"canReadDatabase":false,"canManagePapers":false,"canSubmit":false,"hasHfLevelAccess":false,"upvoted":false,"upvoters":[],"acceptLanguages":["en"],"markdownContentUrl":"https://huggingface.co/buckets/huggingchat/papers-content/resolve/2607/2607.19011.md","query":{}}">
Computational Humor with Multimodal LLMs: Methods, Datasets, Evaluation, and Challenges
Abstract
Multimodal humor in memes, cartoons, and comics remains difficult for AI systems because intended meaning depends on non-literal mechanisms, shared cultural knowledge, and communicative intent rather than literal scene description. This survey focuses on visual humor understanding in single-image and multi-panel artifacts, while treating humor generation as an emerging downstream frontier. We position the literature against prior humor, sarcasm, and general MLLM surveys and organize it using a capability-centric hierarchy spanning recognition, interpretation and reasoning, and generation. Under this lens, we synthesize benchmark design, evaluation protocols, and modeling paradigms, tracing the field's shift from task-specific fusion models to large-model approaches based on multimodal alignment, evidence-grounded reasoning, and controlled generation. We conclude by highlighting the main barriers to progress: shortcut-prone evaluation, limited cultural and narrative coverage, weak evidence grounding, and unresolved safety and ownership concerns.
Community
This survey examines visual humor understanding across memes, cartoons, and comics through a capability-centric hierarchy comprising recognition, interpretation/reasoning, and generation. This taxonomy reorganizes a fragmented literature by focusing on the capabilities that benchmarks and models actually evaluate. Beyond synthesizing prior work, the authors conduct a cross-benchmark evaluation of recent MLLMs, showing that while current models perform reasonably well on visual recognition, they still lag far behind humans on interpretation-intensive tasks. The results also reveal that model rankings vary substantially across humor capabilities and that explicit reasoning variants do not consistently improve performance.
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/2607.19011 in a model README.md to link it from this page.
Cite arxiv.org/abs/2607.19011 in a dataset README.md to link it from this page.
Cite arxiv.org/abs/2607.19011 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.