ROT: Rotating Hidden States towards Contextual Vectors for Hallucination Mitigation in LVLMs

Excerpt

arXiv:2610.06056v1 Announce Type: cross Abstract: Large Vision-Language Models (LVLMs) frequently suffer from object hallucination. Existing training-free interventions primarily manipulate attention weights, which indirectly affect the deep semantics reaching the final predictive layers. In this work, we shift our focus to the hidden state vectors extracted after self-attention and residual addition. Empirical analysis reveals that hallucinated tokens do not simply over-rely on linguistic prior