认知支点与视觉锚定：揭示并纠正多模态推理模型中的幻觉
Cognitive Pivot Points and Visual Anchoring: Unveiling and Rectifying Hallucinations in Multimodal Reasoning Models
AI总结：
针对多模态大推理模型在长链推理中易产生幻觉的问题，提出V-STAR训练范式，通过分层视觉注意力奖励和强制反思机制，将视觉锚定引入推理过程以减轻幻觉。
摘要：
多模态大推理模型（MLRMs）通过测试时计算扩展在视觉推理方面取得了显著进展，但长链推理仍然容易出现幻觉。我们识别出一个称为“推理视觉真相脱节”（RVTD）的令人担忧的现象：幻觉与认知分叉点高度相关，这些分叉点通常表现出高熵状态。我们将这种脆弱性归因于视觉语义锚定的崩溃，这种崩溃位于网络中间层；具体来说，在这些高不确定性过渡期间，模型未能查询视觉证据，而是退回到语言先验。因此，我们主张从仅关注结果层面的监督转向增加细粒度的内部注意力引导。为此，我们提出V-STAR（视觉结构训练与注意力强化），一种轻量级、整体的训练范式，旨在内化视觉感知的推理能力。我们方法的核心是分层视觉注意力奖励（HVAR），集成在GRPO框架内。在检测到高熵状态时，该机制动态激励关键中间层的视觉注意力，从而将推理过程锚定回视觉输入。此外，我们引入了强制反思机制（FRM），一种轨迹编辑策略，通过在高熵认知分叉点触发反思并鼓励后续步骤与视觉输入进行验证，从而打破认知惯性，将外部去偏干预转化为减轻幻觉的内在能力。
英文摘要：
Multimodal Large Reasoning Models (MLRMs) have achieved remarkable strides in visual reasoning through test time compute scaling, yet long chain reasoning remains prone to hallucinations. We identify a concerning phenomenon termed the Reasoning Vision Truth Disconnect (RVTD): hallucinations are strongly correlated with cognitive bifurcation points that often exhibit high entropy states. We attribute this vulnerability to a breakdown in visual semantic anchoring, localized within the network's intermediate layers; specifically, during these high uncertainty transitions, the model fails to query visual evidence, reverting instead to language priors. Consequently, we advocate a shift from solely outcome level supervision to augmenting it with fine grained internal attention guidance. To this end, we propose V-STAR (Visual Structural Training with Attention Reinforcement), a lightweight, holistic training paradigm designed to internalize visually aware reasoning capabilities. Central to our approach is the Hierarchical Visual Attention Reward (HVAR), integrated within the GRPO framework. Upon detecting high entropy states, this mechanism dynamically incentivizes visual attention across critical intermediate layers, thereby anchoring the reasoning process back to the visual input. Furthermore, we introduce the Forced Reflection Mechanism (FRM), a trajectory editing strategy that disrupts cognitive inertia by triggering reflection around high entropy cognitive bifurcation points and encouraging verification of subsequent steps against the visual input, thereby translating external debiasing interventions into an intrinsic capability for hallucination mitigation.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
