mlsanityData to 14 Oct 2026

MemeBench: What LVLMs Miss When Interpreting Culture-Dependent Memes

arXiv 2607.27798 ↗First posted 30 Jul 2026cs.AIcited by 1

On the map: Multimodal LLMs / vision-language models → Document understanding and multimodal RAG

Abstract (arXiv)

Large vision-language models have improved at describing visual content, but accurate descriptions do not ensure interpretation when meaning depends on knowledge beyond the pixels. Memes expose this gap because they rely on cultural entities, background knowledge, and community conventions. Most meme benchmarks reduce interpretation to labels or holistic scores, obscuring where an explanation breaks down. We introduce MemeBench, a diagnostic benchmark of 1,253 Chinese and English memes with human-written references and quality-controlled VIKR annotations, centered on anime, comics, games, and adjacent online subcultures. Its VIKR schema decomposes explanations into Visual clues, Identity links, Knowledge units, and Reasoning mechanisms. Across 26 LVLMs, every model covers visible content more reliably than the knowledge needed to interpret it, and even the strongest retains a 22.6% Visual-Knowledge gap. To test whether this diagnosis can guide improvement, we introduce KAR, an entity-guided retrieval baseline built on CultureBase. Across four controlled models, KAR raises VIKR Success by 3.6-7.4% and, compared with generic retrieval, repairs more answers and breaks fewer. Yet both retrieval conditions improve Identity and Knowledge while reducing Visual coverage in every comparison. MemeBench reveals whether an interpretation succeeds, what is missing, and whether targeted evidence fills the diagnosed gap.

Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Placed in Document understanding and multimodal RAG, one of 16 directions in Multimodal LLMs / vision-language models. Builds on 25 of its 40 references that are in the map. First posted on arXiv 30 Jul 2026.

Builds on

25 of 40 references.

Closest ancestors · 3

Same direction, newest first

  1. 2025MMSearch-R1: Incentivizing LMMs to Search25 Jun
  2. 2024SearchLVLMs: A Plug-and-Play Framework for Augmenting Large Vision-Language Models by Searching Up-to-Date Internet Knowledge23 May
  3. 2024Wiki-LLaVA: Hierarchical Retrieval-Augmented Generation for Multimodal LLMs23 Apr

Other references · 22

  1. 2020Retrieval-Augmented Generation for Knowledge-Intensive NLP Taskscited by 9,297
  2. 2023Judging LLM-as-a-Judge with MT-Bench and Chatbot Arenacited by 8,626
  3. 2025Qwen3-VL Technical Reportcited by 2,793
  4. 2023MMBench: Is Your Multi-modal Model an All-around Player?cited by 2,141
  5. 2023MMMU: A Massive Multi-discipline Multimodal Understanding and Reasoning Benchmark for Expert AGIcited by 2,080
  6. 2023MM-Vet: Evaluating Large Multimodal Models for Integrated Capabilitiescited by 1,239
  7. 2019OK-VQA: A Visual Question Answering Benchmark Requiring External Knowledgecited by 1,121
  8. 2020The Hateful Memes Challenge: Detecting Hate Speech in Multimodal Memescited by 490
  9. 2022MuRAG: Multimodal Retrieval-Augmented Generator for Open Question Answering over Images and Textcited by 220
  10. 2023Can Pre-trained Vision and Language Models Answer Visual Information-Seeking Questions?cited by 207
  11. 2024BLEnD: A Benchmark for LLMs on Everyday Knowledge in Diverse Cultures and Languagescited by 173
  12. 2023Encyclopedic VQA: Visual questions about detailed properties of fine-grained categoriescited by 107
10 more
  1. 2024Benchmarking Vision Language Models for Cultural Understandingcited by 101
  2. 2024CheckEval: A reliable LLM-as-a-Judge framework for evaluating text generation using checklistscited by 94
  3. 2021MOMENTA: A Multimodal Framework for Detecting Harmful Memes and Their Targetscited by 86
  4. 2024Can MLLMs Understand the Deep Implication Behind Chinese Images?cited by 14
  5. 2024Towards Comprehensive Detection of Chinese Harmful Memescited by 13
  6. 2025D-HUMOR: Dark Humor Understanding via Multimodal Open-ended Reasoning -- A Benchmark Dataset and Methodcited by 7
  7. 2026MemeLens: Multilingual Multitask VLMs for Memescited by 6
  8. 2026AVMeme Exam: A Multimodal Multilingual Multicultural Benchmark for LLMs' Contextual and Cultural Knowledge and Thinkingcited by 5
  9. 2026M-QUEST -- Meme Question-Understanding Evaluation on Semantics and Toxicitycited by 3
  10. 2026I Came, I Saw, I Explained: Benchmarking Multimodal LLMs on Figurative Meaning in Memescited by 2

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 1.

1 paper.

  1. 2026MemeMind: Reference-Guided Trace Construction for Offline Context Optimizationcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026MemeCULT-1K: Benchmarking South Asian Cultural Context and Humor Understanding of Multimodal Models
  2. 2026Order Matters: A Chinese Multi-Panel Meme Benchmark for Vision-Language Reasoning
  3. 2025MemeReaCon: Probing Contextual Meme Understanding in Large Vision-Language Models
  4. 2026Beyond Translation: Cross-Cultural Meme Transcreation with Vision-Language Models
  5. 2025Large Vision-Language Models for Knowledge-Grounded Data Annotation of Memes
  6. 2026BanglaMemeX: Advancing Cultural Metaphoric Image Interpretation in Bangla with a Multimodal Explainable Dataset
  7. 2023MemeCap: A Dataset for Captioning and Interpreting Memes
  8. 2026They Said Memes Were Harmless-We Found the Ones That Hurt: Decoding Jokes, Symbols, and Cultural References