# 通过阶段性自奖励缓解多模态幻觉

- 来源：HuggingFace Daily Papers（社区热门论文）
- 发布时间：2026-04-20 08:00
- AIHOT 链接：https://aihot.virxact.com/items/cmo9ps1nz03vjsls2rgs27xlh
- 原文链接：https://arxiv.org/abs/2604.17982

## AI 摘要

研究团队提出PSRD（分阶段自奖励解码）框架，在推理阶段动态缓解大型视觉语言模型（LVLMs）的视觉幻觉问题，无需外部监督。研究发现幻觉在每个语义阶段开始时达到峰值，据此将LVLMs的幻觉引导信号蒸馏为轻量级奖励模型，实现解码过程中的实时干预。实验显示，该方法使LLaVA-1.5-7B的幻觉率降低50.0%，并在五个幻觉评估基准上持续优于现有事后方法，同时实现了性能与推理效率的可控平衡。

## 正文

Large Vision-Language Models (LVLMs) still struggle with vision hallucination, where generated responses are inconsistent with the visual input. Existing methods either rely on large-scale annotated data for fine-tuning, which incurs massive computational overhead, or employ static post-hoc strategies that overlook the dynamic nature of hallucination emergence. To address these, we introduce a new self-rewarding framework, enabling dynamic hallucination mitigation at inference time without external supervision. On the empirical side, we reveal that visual hallucination exhibits phase-wise dynamic patterns, peaking at the onset of each semantic phase. Drawing on these insights, we propose PSRD (Phase-wise \textbf{Self-Reward Decoding) for online hallucination correction guided by phase-wise self-reward signals. To reduce the cost of repeated self-evaluation during decoding, we distill the hallucination guidance signal from LVLMs into a lightweight reward model. The reward model subsequently provides on-the-fly guidance for targeted intervention during the decoding process, enabling precise hallucination suppression. The proposed PSRD significantly reduces the hallucination rate of LLaVA-1.5-7B by 50.0% and consistently outperforms existing post-hoc methods across five hallucination evaluation benchmarks for four LVLMs. Further analysis confirms that PSRD effectively mitigates hallucination propagation and achieves a highly controllable trade-off between strong performance and inference efficiency.