mlsanityData to 14 Oct 2026

Can Large Vision-Language Models Understand Multimodal Sarcasm?

arXiv 2508.03654 ↗First posted 5 Aug 2025cs.CL cs.CVcited by 5

On the map: Multimodal LLMs / vision-language models → MLLM training and visual instruction tuning

Abstract (arXiv)

Sarcasm is a complex linguistic phenomenon that involves a disparity between literal and intended meanings, making it challenging for sentiment analysis and other emotion-sensitive tasks. While traditional sarcasm detection methods primarily focus on text, recent approaches have incorporated multimodal information. However, the application of Large Visual Language Models (LVLMs) in Multimodal Sarcasm Analysis (MSA) remains underexplored. In this paper, we evaluate LVLMs in MSA tasks, specifically focusing on Multimodal Sarcasm Detection and Multimodal Sarcasm Explanation. Through comprehensive experiments, we identify key limitations, such as insufficient visual understanding and a lack of conceptual knowledge. To address these issues, we propose a training-free framework that integrates in-depth object extraction and external conceptual knowledge to improve the model's ability to interpret and explain sarcasm in multimodal contexts. The experimental results on multiple models show the effectiveness of our proposed framework. The code is available at https://github.com/cp-cp/LVLM-MSA.

Key facts. Cited by 5 papers in the map (data through 14 Oct 2026). Placed in MLLM training and visual instruction tuning, one of 16 directions in Multimodal LLMs / vision-language models. Builds on 23 of its 51 references that are in the map. First posted on arXiv 5 Aug 2025.

Builds on

23 of 51 references.

Closest ancestors · 3

Same direction, newest first

  1. 2024MM-InstructEval: Zero-Shot Evaluation of (Multimodal) Large Language Models on Multimodal Reasoning Tasks12 May
  2. 2023Beyond Task Performance: Evaluating and Reducing the Flaws of Large Multimodal Models with In-Context Learning1 Oct
  3. 2023InstructBLIP: Towards General-purpose Vision-Language Models with Instruction Tuning11 May

Other references · 20

  1. 2023LLaMA: Open and Efficient Foundation Language Modelscited by 14,132
  2. 2015Fast R-CNNcited by 2,933
  3. 2016ConceptNet 5.5: An Open Multilingual Graph of General Knowledgecited by 1,067
  4. 2023Multimodal Chain-of-Thought Reasoning in Language Modelscited by 752
  5. 2023DDCoT: Duty-Distinct Chain-of-Thought Prompting for Multimodal Reasoning in Language Modelscited by 128
  6. 2024GOAT-Bench: Safety Insights to Large Multimodal Models through Meme-Based Social Abusecited by 52
  7. 2024LEMMA: Towards LVLM-Enhanced Multimodal Misinformation Detection with External Knowledge Augmentationcited by 28
  8. 2016Modelling Context with User Embeddings for Sarcasm Detection in Social Mediacited by 26
  9. 2021Nice perfume. How long did you marinate in it? Multimodal Sarcasm Explanationcited by 25
  10. 2023Multi-source Semantic Graph-based Multimodal Sarcasm Explanation Generationcited by 19
  11. 2022Counterfactual Reasoning for Out-of-distribution Multimodal Sentiment Analysiscited by 18
  12. 2024WisdoM: Improving Multimodal Sentiment Analysis by Fusing Contextual World Knowledgecited by 16
8 more
  1. 2016Are Word Embedding-based Features Useful for Sarcasm Detection?cited by 14
  2. 2023Stylized Data-to-Text Generation: A Case Study in the E-Commerce Domaincited by 11
  3. 2023General Debiasing for Multimodal Sentiment Analysiscited by 10
  4. 2024Fine-grained and Explainable Factuality Evaluation for Multimodal Summarizationcited by 10
  5. 2024CofiPara: A Coarse-to-fine Paradigm for Multimodal Sarcasm Target Identification with Large Multimodal Modelscited by 8
  6. 2024Sentiment-enhanced Graph-based Sarcasm Explanation in Dialoguecited by 6
  7. 2025Rainbow Noise: Stress-Testing Multimodal Harmful-Meme Detectors on LGBTQ Contentcited by 6
  8. 2024Defeasible Visual Entailment: Benchmark, Evaluator, and Reward-Driven Optimizationcited by 3

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 5.

5 papers.

  1. 2026"I See What You Did There": Can Large Vision-Language Models Understand Multimodal Puns?cited by 2
  2. 2026Reasoning Beyond Literal: Cross-style Multimodal Reasoning for Figurative Language Understandingcited by 0
  3. 2026GRASP: Grounded CoT Reasoning with Dual-Stage Optimization for Multimodal Sarcasm Target Identificationcited by 0
  4. 2026MetaphorVU: Towards Metaphorical Video Understandingcited by 0
  5. 2026Beyond Visual Similarity: Entity-Aligned Retrieval for Knowledge-Based Visual Question Answeringcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025Evaluating Open-Source Vision-Language Models for Multimodal Sarcasm Detection
  2. 2025Seeing Sarcasm Through Different Eyes: Analyzing Multimodal Sarcasm Perception in Large Vision-Language Models
  3. 2025Evaluating Multimodal Large Language Models on Spoken Sarcasm Understanding
  4. 2024SarcasmBench: Towards Evaluating Large Language Models on Sarcasm Understanding
  5. 2025MuSaG: A Multimodal German Sarcasm Dataset with Full-Modal Annotations
  6. 2025Think Twice Before You Judge: Mixture of Dual Reasoning Experts for Multimodal Sarcasm Detection
  7. 2025MMSD3.0: A Multi-Image Benchmark for Real-World Multimodal Sarcasm Detection
  8. 2025MUStReason: A Benchmark for Diagnosing Pragmatic Reasoning in Video-LMs for Multimodal Sarcasm Detection