SOURCE-LINKED INTELLIGENCE

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

arXiv · AI, language, vision and robotics · article · Aug 29, 2026 · UTC

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 generatio

Read original source ↗ Open in workspace

recordType

paper

region

Global