mlsanityData to 7 Oct 2026

EviAnchor: Mitigating Hallucinations in Large Vision-Language Models via Regional Visual Evidence Compensation

arXiv 2608.29092 ↗First posted 29 Aug 2026cs.AI cs.CVcited by 0

On the map: Multimodal LLMs / vision-language models → VLM benchmarks and visual reasoning

Abstract (arXiv)

Large vision-language models (LVLMs) frequently generate content unsupported by visual inputs. Preliminary experiments show that visual evidence is primarily incorporated into answer-side representations in early-to-middle decoder layers, while its direct influence progressively weakens in later layers. This attenuation suggests that visual evidence acquired earlier may be insufficiently utilized during subsequent generation. Based on this observation, we propose EviAnchor, a training-free and single-branch inference framework that preserves and reactivates visual evidence throughout generation. EviAnchor introduces Regional Evidence Anchor (REA) slots to progressively aggregate dense visual tokens into spatially structured representations. It then strengthens the current decision state's access to these visual anchors through decision-conditioned evidence routing, mitigating excessive dependence on textual context. Finally, the model resumes its native Transformer computation to integrate the retrieved visual evidence with question semantics and generation history. Experiments across POPE, CHAIR, and MMHal-Bench demonstrate consistent improvements in visual grounding.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in VLM benchmarks and visual reasoning, one of 16 directions in Multimodal LLMs / vision-language models. Builds on 18 of its 27 references that are in the map. First posted on arXiv 29 Aug 2026.

Builds on

18 of 27 references.

References · 18

  1. 2023InstructBLIP: Towards General-purpose Vision-Language Models with Instruction Tuningcited by 2,855
  2. 2023Aligning Large Multimodal Models with Factually Augmented RLHFcited by 703
  3. 2023Mitigating Object Hallucinations in Large Vision-Language Models through Visual Contrastive Decodingcited by 671
  4. 2018Object Hallucination in Image Captioningcited by 661
  5. 2023OPERA: Alleviating Hallucination in Multi-Modal Large Language Models via Over-Trust Penalty and Retrospection-Allocationcited by 468
  6. 2023Mitigating Hallucination in Large Multi-Modal Models via Robust Instruction Tuningcited by 439
  7. 2023DoLa: Decoding by Contrasting Layers Improves Factuality in Large Language Modelscited by 367
  8. 2024Mitigating Hallucinations in Large Vision-Language Models with Instruction Contrastive Decodingcited by 198
  9. 2024HALC: Object Hallucination Reduction via Adaptive Focal-Contrast Decodingcited by 171
  10. 2024Devils in Middle Layers of Large Vision-Language Models: Interpreting, Detecting and Mitigating Object Hallucinations via Attention Lenscited by 139
  11. 2024IBD: Alleviating Hallucinations in Large Vision-Language Models via Image-Biased Decodingcited by 118
  12. 2024Nullu: Mitigating Object Hallucinations in Large Vision-Language Models via HalluSpace Projectioncited by 46
6 more
  1. 2025Seeing Far and Clearly: Mitigating Hallucinations in MLLMs with Attention Causal Decodingcited by 42
  2. 2026Vision-Language Introspection: Mitigating Overconfident Hallucinations in MLLMs via Interpretable Bi-Causal Steeringcited by 8
  3. 2024Prescribing the Right Remedy: Mitigating Hallucinations in Large Vision-Language Models via Targeted Instruction Tuningcited by 6
  4. 2025Mitigating Hallucinations in Large Vision-Language Models by Self-Injecting Hallucinationscited by 5
  5. 2026Where Does Vision Meet Language? Understanding and Refining Visual Fusion in MLLMs via Contrastive Attentioncited by 3
  6. 2026Decoding by Perturbation: Mitigating MLLM Hallucinations via Dynamic Textual Perturbationcited by 3

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026Look Where It Counts: A Free, Label-Free Visual Evidence Signal for Fine-Grained Vision-Language Reasoning
  2. 2025VisRAG2.0: Mitigating Visual Hallucinations via Evidence-Guided Multi-Image Reasoning in Visual Retrieval-Augmented Generation
  3. 2025Seeing but Not Believing: Probing the Disconnect Between Visual Attention and Answer Correctness in VLMs
  4. 2026Seeing is not Enough: Vision-Language Models Perceive Evidence but Fail to Act
  5. 2026Diagnosing Visual Ignorance in Vision-Language Models
  6. 2026DeepScan: A Training-Free Framework for Visually Grounded Reasoning in Large Vision-Language Models
  7. 2024FiVL: A Framework for Improved Vision-Language Alignment through the Lens of Training, Evaluation and Explainability
  8. 2026CDH-Bench: A Commonsense-Driven Hallucination Benchmark for Evaluating Visual Fidelity in Vision-Language Models