mlsanityData to 14 Oct 2026

Online Self-Calibration Against Hallucination in Vision-Language Models

arXiv 2605.00323 ↗First posted 1 May 2026cs.CV cs.LGcited by 0

On the map: Multimodal LLMs / vision-language models → Hallucination mitigation in LVLMs

Abstract (arXiv)

Large Vision-Language Models (LVLMs) often suffer from hallucinations, generating descriptions that include visual details absent from the input image. Recent preference alignment methods typically rely on supervision distilled from stronger models such as GPT. However, this offline paradigm introduces a Supervision-Perception Mismatch: the student model is forced to align with fine-grained details beyond its perceptual capacity, learning to guess rather than to see. To obtain reliable self-supervision for online learning, we identify a Generative-Discriminative Gap within LVLMs, where models exhibit higher accuracy on discriminative verification than open-ended generation. Leveraging this capability, we propose Online Self-CAlibRation (OSCAR), a framework that integrates Monte Carlo Tree Search with a Dual-Granularity Reward Mechanism to construct preference data and iteratively refines the model via Direct Preference Optimization. Extensive experiments demonstrate that OSCAR achieves state-of-the-art performance on hallucination benchmarks while improving general multimodal capabilities.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Hallucination mitigation in LVLMs, one of 16 directions in Multimodal LLMs / vision-language models. Builds on 47 of its 52 references that are in the map. First posted on arXiv 1 May 2026.

Builds on

47 of 52 references.

Closest ancestors · 11

Same direction, newest first

  1. 2025Octopus: Alleviating Hallucination via Dynamic Contrastive Decoding1 Mar
  2. 2024V-DPO: Mitigating Hallucination in Large Vision Language Models via Vision-Guided Direct Preference Optimization5 Nov
  3. 2024Hallucination of Multimodal Large Language Models: A Survey29 Apr
  4. 2024A Survey on Hallucination in Large Vision-Language Models1 Feb
  5. 2023OPERA: Alleviating Hallucination in Multi-Modal Large Language Models via Over-Trust Penalty and Retrospection-Allocation29 Nov
  6. 2023Mitigating Object Hallucinations in Large Vision-Language Models through Visual Contrastive Decoding28 Nov
  7. 2023AMBER: An LLM-free Multi-dimensional Benchmark for MLLMs Hallucination Evaluation13 Nov
  8. 2023Woodpecker: Hallucination Correction for Multimodal Large Language Models24 Oct
  9. 2023Detecting and Preventing Hallucinations in Large Vision Language Models11 Aug
  10. 2023Mitigating Hallucination in Large Multi-Modal Models via Robust Instruction Tuning26 Jun
  11. 2023Evaluating Object Hallucination in Large Vision-Language Models17 May

Other references · 36

  1. 2023GPT-4 Technical Reportcited by 18,396
  2. 2025Qwen3 Technical Reportcited by 8,820
  3. 2023Direct Preference Optimization: Your Language Model is Secretly a Reward Modelcited by 8,065
  4. 2023Visual Instruction Tuningcited by 7,869
  5. 2023Qwen Technical Reportcited by 3,344
  6. 2023InstructBLIP: Towards General-purpose Vision-Language Models with Instruction Tuningcited by 2,876
  7. 2023MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large Language Modelscited by 2,711
  8. 2023MM-Vet: Evaluating Large Multimodal Models for Integrated Capabilitiescited by 1,239
  9. 2023Reasoning with Language Model is Planning with World Modelcited by 861
  10. 2024DeepSeek-VL: Towards Real-World Vision-Language Understandingcited by 733
  11. 2023Aligning Large Multimodal Models with Factually Augmented RLHFcited by 712
  12. 2024Self-Rewarding Language Modelscited by 681
24 more
  1. 2018Object Hallucination in Image Captioningcited by 669
  2. 2022Large Language Models Can Self-Improvecited by 664
  3. 2023mPLUG-Owl2: Revolutionizing Multi-modal Large Language Model with Modality Collaborationcited by 547
  4. 2023A Survey of Hallucination in Large Foundation Modelscited by 368
  5. 2024Monte Carlo Tree Search Boosts Reasoning via Iterative Preference Learningcited by 241
  6. 2025Dynamic Early Exit in Reasoning Modelscited by 219
  7. 2024Aligning Modalities in Vision Large Language Models via Preference Fine-tuningcited by 198
  8. 2024Accessing GPT-4 level Mathematical Olympiad Solutions via Monte Carlo Tree Self-refine with LLaMa-3 8Bcited by 158
  9. 2023Silkie: Preference Distillation for Large Visual Language Modelscited by 130
  10. 2024Toward Self-Improvement of LLMs via Imagination, Searching, and Criticizingcited by 125
  11. 2025S-GRPO: Early Exit via Reinforcement Learning in Reasoning Modelscited by 87
  12. 2024Enhancing Visual-Language Modality Alignment in Large Vision Language Models via Self-Improvementcited by 74
  13. 2024Enhancing Large Vision Language Models with Self-Training on Image Comprehensioncited by 70
  14. 2022Language Prior Is Not the Only Shortcut: A Benchmark for Shortcut Learning in VQAcited by 30
  15. 2021Check It Again: Progressive Visual Question Answering via Visual Entailmentcited by 18
  16. 2024ChatSOP: An SOP-Guided MCTS Planning Framework for Controllable LLM Dialogue Agentscited by 12
  17. 2022Towards Robust Visual Question Answering: Making the Most of Biased Samples via Contrastive Learningcited by 9
  18. 2023Combo of Thinking and Observing for Outside-Knowledge VQAcited by 9
  19. 2024Teaching Language Models to Self-Improve by Learning from Language Feedbackcited by 9
  20. 2025The Self-Improvement Paradox: Can Language Models Bootstrap Reasoning Capabilities without External Scaffolding?cited by 7
  21. 2024Think&Cite: Improving Attributed Text Generation with Self-Guided Tree Search and Progress Reward Modelingcited by 6
  22. 2025Weights-Rotated Preference Optimization for Large Language Modelscited by 3
  23. 2026System 1&2 Synergy via Dynamic Model Interpolationcited by 3
  24. 2025Breaking the Trade-Off Between Faithfulness and Expressiveness for Large Language Modelscited by 2

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2025OViP: Online Vision-Language Preference Learning for VLM Hallucination
  2. 2025Mitigating Hallucinations in Large Vision-Language Models by Self-Injecting Hallucinations
  3. 2026Aligning with Your Own Voice: Self-Corrected Preference Learning for Hallucination Mitigation in LVLMs
  4. 2024Token Preference Optimization with Self-Calibrated Visual-Anchored Rewards for Hallucination Mitigation
  5. 2026Grounded Post-Training with Hard Examples for Reducing Hallucination in Multimodal Large Language Models
  6. 2024Cracking the Code of Hallucination in LVLMs with Vision-aware Head Divergence
  7. 2025Self-Consistency as a Free Lunch: Reducing Hallucinations in Vision-Language Models via Self-Reflection
  8. 2025Mitigating Hallucinations in Large Vision-Language Models via Entity-Centric Multimodal Preference Optimization