mlsanityData to 14 Oct 2026

Insight Over Sight: Exploring the Vision-Knowledge Conflicts in Multimodal LLMs

arXiv 2410.08145 ↗First posted 10 Oct 2024cs.CL cs.CVcited by 38

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

Abstract (arXiv)

This paper explores the problem of commonsense level vision-knowledge conflict in Multimodal Large Language Models (MLLMs), where visual information contradicts model's internal commonsense knowledge. To study this issue, we introduce an automated framework, augmented with human-in-the-loop quality control, to generate inputs designed to simulate and evaluate these conflicts in MLLMs. Using this framework, we have crafted a diagnostic benchmark consisting of 374 original images and 1,122 high-quality question-answer (QA) pairs. The benchmark covers two aspects of conflict and three question types, providing a thorough assessment tool. We apply this benchmark to assess the conflict-resolution capabilities of nine representative MLLMs from various model families. Our results indicate an evident over-reliance on parametric knowledge for approximately 20% of all queries, especially among Yes-No and action-related problems. Based on these findings, we evaluate the effectiveness of existing approaches to mitigating the conflicts and compare them to our "Focus-on-Vision" prompting strategy. Despite some improvement, the vision-knowledge conflict remains unresolved and can be further scaled through our data construction framework. Our proposed framework, benchmark, and analysis contribute to the understanding and mitigation of vision-knowledge conflicts in MLLMs.

Key facts. Cited by 38 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 27 of its 47 references that are in the map. First posted on arXiv 10 Oct 2024.

Builds on

27 of 47 references.

Closest ancestors · 4

Same direction, newest first

  1. 2024Cambrian-1: A Fully Open, Vision-Centric Exploration of Multimodal LLMs24 Jun
  2. 2023Improved Baselines with Visual Instruction Tuning5 Oct
  3. 2023Macaw-LLM: Multi-Modal Language Modeling with Image, Audio, Video, and Text Integration15 Jun
  4. 2023InstructBLIP: Towards General-purpose Vision-Language Models with Instruction Tuning11 May

Other references · 23

  1. 2020Language Models are Few-Shot Learnerscited by 32,419
  2. 2023Llama 2: Open Foundation and Fine-Tuned Chat Modelscited by 13,028
  3. 2021Finetuned Language Models Are Zero-Shot Learnerscited by 3,780
  4. 2015VQA: Visual Question Answeringcited by 2,506
  5. 2023Qwen-VL: A Versatile Vision-Language Model for Understanding, Localization, Text Reading, and Beyondcited by 2,224
  6. 2023MMMU: A Massive Multi-discipline Multimodal Understanding and Reasoning Benchmark for Expert AGIcited by 2,080
  7. 2015Microsoft COCO Captions: Data Collection and Evaluation Servercited by 1,462
  8. 2016CLEVR: A Diagnostic Dataset for Compositional Language and Elementary Visual Reasoningcited by 1,457
  9. 2023Mitigating Object Hallucinations in Large Vision-Language Models through Visual Contrastive Decodingcited by 688
  10. 2023OPERA: Alleviating Hallucination in Multi-Modal Large Language Models via Over-Trust Penalty and Retrospection-Allocationcited by 478
  11. 2023Mitigating Hallucination in Large Multi-Modal Models via Robust Instruction Tuningcited by 445
  12. 2021Entity-Based Knowledge Conflicts in Question Answeringcited by 356
11 more
  1. 2018AllenNLP: A Deep Semantic Natural Language Processing Platformcited by 347
  2. 2023Adaptive Chameleon or Stubborn Sloth: Revealing the Behavior of Large Language Models in Knowledge Conflictscited by 341
  3. 2023AMBER: An LLM-free Multi-dimensional Benchmark for MLLMs Hallucination Evaluationcited by 300
  4. 2024Knowledge Conflicts for LLMs: A Surveycited by 283
  5. 2024Paying More Attention to Image: A Training-Free Method for Alleviating Hallucination in LVLMscited by 179
  6. 2024Large Language Models are Inconsistent and Biased Evaluatorscited by 153
  7. 2022Rich Knowledge Sources Bring Complex Knowledge Conflicts: Recalibrating Models to Reflect Conflicting Evidencecited by 144
  8. 2015Grounding of Textual Phrases in Images by Reconstructioncited by 112
  9. 2024ConflictBank: A Benchmark for Evaluating the Influence of Knowledge Conflicts in LLMcited by 73
  10. 2023Beyond Factuality: A Comprehensive Evaluation of Large Language Models as Knowledge Generatorscited by 43
  11. 2021On Compositional Generalization of Neural Machine Translationcited by 34

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2024: 2; 2025: 12; 2026: 24.

38 papers.

Sort:

By area:

  1. 2025MLLMs are Deeply Affected by Modality Biascited by 40
  2. 2025Evaluating and Steering Modality Preferences in Multimodal Large Language Modelcited by 33
  3. 2026STEP3-VL-10B Technical Reportcited by 32
  4. 2025When Modalities Conflict: How Unimodal Reasoning Uncertainty Governs Preference Dynamics in MLLMscited by 21
  5. 2025Benchmarking Multimodal Knowledge Conflict for Large Multimodal Modelscited by 19
  6. 2025Robust Multimodal Large Language Models Against Modality Conflictcited by 18
  7. 2025Task Matters: Knowledge Requirements Shape LLM Responses to Context-Memory Conflictcited by 16
  8. 2025VISTA: Enhancing Vision-Text Alignment in MLLMs via Cross-Modal Mutual Information Maximizationcited by 12
  9. 2026DM0: An Embodied-Native Vision-Language-Action Model towards Physical AIcited by 12
  10. 2025When Seeing Overrides Knowing: Disentangling Knowledge Conflicts in Vision-Language Modelscited by 11
  11. 2024Hallucination in Multimodal Foundation Models: A Survey on Causes, Corrections, and Evaluationscited by 8
  12. 2025Mitigating Modal Imbalance in Multimodal Reasoningcited by 8
26 more
  1. 2026V-FAT: Benchmarking Visual Fidelity Against Text-biascited by 6
  2. 2025Segmentation as A Plug-and-Play Capability for Frozen Multimodal LLMscited by 5
  3. 2024Is Cognition Consistent with Perception? Assessing and Mitigating Multimodal Knowledge Conflicts in Document Understandingcited by 4
  4. 2025Unleashing the Intrinsic Visual Representation Capability of Multimodal Large Language Modelscited by 2
  5. 2025What Color Is the Text? A Benchmark for Hallucination Induced by Image-Embedded Promptcited by 1
  6. 2026When RAG Hurts: Diagnosing and Mitigating Attention Distraction in Retrieval-Augmented LVLMscited by 1
  7. 2026Fix Before Search: Benchmarking Agentic Query Visual Pre-processing in Multimodal Retrieval-augmented Generationcited by 1
  8. 2026REAL: Resolving Knowledge Conflicts in Knowledge-Intensive Visual Question Answering via Reasoning-Pivot Alignmentcited by 1
  9. 2026Diagnosing Knowledge Conflict in Multimodal Long-Chain Reasoningcited by 1
  10. 2026Measuring Cross-Modal Synergy: A Benchmark for VLM Explainabilitycited by 1
  11. 2026Do Multimodal LLMs See Before They Read? Diagnosing Contextual Sycophancycited by 1
  12. 2026Slow to See, Slow to Suppress: Understanding the Effects of Modality in Context-Memory Conflictscited by 1
  13. 2026Seeing is not Enough: Vision-Language Models Perceive Evidence but Fail to Actcited by 1
  14. 2026MLLMs Get It Right, Then Get It Wrong: Tracing and Correcting Late-Layer Textual Biascited by 0
  15. 2026Multimodal Knowledge Edit-Scoped Generalization for Online Recursive MLLM Editingcited by 0
  16. 2026MissingBench-Verified: Probing Vision-Language Models' Inability to Detect Missing Object Partscited by 0
  17. 2026ENTRAP-VL: A Taxonomic Probe for Dual Contextual Entrainment in Vision-Language Modelscited by 0
  18. 2026CAER: Conflict-Aware Evidence Routing with Dual Prefix Experts for Multimodal Large Language Modelscited by 0
  19. 2026When Model Priors Conflict with Visual Evidence: Mitigating Commonsense-Driven Hallucinations by Selective Prior Calibrationcited by 0
  20. 2026GeoArbiter: Verifiability-Guided Grounding for Remote-Sensing Multimodal LLMscited by 0
  21. 2026When Text and Numbers Disagree: Evidence Arbitration in Large Language Modelscited by 0
  22. 2026Same evidence, different judgments: Evidence noncommutative in vision/speech-text conflictscited by 0
  23. 2026When Helpful Text Hurts: Option-Redirecting Bias in Vision-Language Modelscited by 0
  24. 2026When Words Speak Louder than Images: Towards Understanding Language Bias in Vision-Language Modelscited by 0
  25. 2026Selective Listening: Mechanism-Guided Control of Audio Influence in Large Audio-Language Modelscited by 0
  26. 2026Vision-Language Models Integrate Conflicting Evidence Like Biased Bayesian Observerscited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2024Unraveling Cross-Modality Knowledge Conflicts in Large Vision-Language Models
  2. 2026CC-VQA: Conflict- and Correlation-Aware Method for Mitigating Knowledge Conflict in Knowledge-Based Visual Question Answering
  3. 2026Same Semantics, Different Outcome: On the Modality Robustness of Multimodal LLMs under Knowledge Conflict
  4. 2024Understanding the Role of LLMs in Multimodal Evaluation Benchmarks
  5. 2025Debating for Better Reasoning: An Unsupervised Multimodal Approach
  6. 2024Evaluating and Advancing Multimodal Large Language Models in Perception Ability Lens
  7. 2024MMEvalPro: Calibrating Multimodal Benchmarks Towards Trustworthy and Efficient Evaluation
  8. 2025Evaluating Variance in Visual Question Answering Benchmarks