Entity · paper

Computational Humor with Multimodal LLMs: Methods, Datasets, Evaluation, and Challenges

paperactiveprovisionalcomputational-humor-with-multimodal-llms-methods-datasets-evaluation-and-challenges-97813955·1 events·first seen Jul 22, 2026

Aliases: Computational Humor with Multimodal LLMs: Methods, Datasets, Evaluation, and Challenges

More like this (12)

Multimodal Large Language Models Benchmarking Multimodal Large Language Models for Scientific Visualization Literacy Multimodal Learning Where Do Models Find Happiness? Emotion Vectors in Open-Source LLMs Multimodal Augmented Generation via Multimodal Retrieval Workshop Unified Multimodal Models (UMMs)ICML 2026 Workshop on Efficient Multimodal Question Answering multimodal classification models Latent World Recovery for Multimodal Learning with Missing Modalities multimodal embedding MODUS: Decoder-Only Any-to-Any Modeling of Diverse Modalities Data Journalist Agent: Transforming Data into Verifiable Multimodal Stories

Recent events (1)

4arXiv · cs.CL·Jul 22, 2026·source ↗

Survey on computational humor with multimodal LLMs: benchmarks, methods, and open challenges

A new arXiv survey examines the state of visual humor understanding in AI systems, covering memes, cartoons, and comics as test cases for non-literal reasoning. The authors organize the literature into a capability hierarchy spanning recognition, interpretation, reasoning, and generation, and trace the field's shift from task-specific fusion models to large multimodal model approaches. Key barriers identified include shortcut-prone evaluation, limited cultural coverage, weak evidence grounding, and safety and ownership concerns.

Evaluation and Benchmarking Multimodal Progress Computational Humor with Multimodal LLMs: Methods, Datasets, Evaluation, and Challenges