Cognitive Pivot Points and Visual Anchoring: Unveiling and Rectifying Hallucinations in Multimodal Reasoning Models

• URL: http://arxiv.org/abs/2604.10219v1
• Date: Sat, 11 Apr 2026 13:59:05 GMT
• Title: Cognitive Pivot Points and Visual Anchoring: Unveiling and Rectifying Hallucinations in Multimodal Reasoning Models
• Abstract summary: hallucinations are strongly correlated with cognitive bifurcation points that often exhibit high entropy states.<n>We propose V-STAR, a lightweight, holistic training paradigm designed to internalize visually aware reasoning capabilities.
• Score: 65.57985131861399
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: 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.

Related papers

• See Only When Needed: Context-Aware Attention Intervention for Mitigating Hallucinations in LVLMs [81.71384150441163]
  We propose Context-aware Attention Intervention (CAI) for large vision-language models (LVLMs)<n>CAI enforces a see only when needed principle via two-axis selectivity: where to look and when to intervene.<n> experiments across multiple LVLM backbones and benchmarks show that CAI achieves state-of-the-art hallucination mitigation.
  arXiv  Detail & Related papers  (2026-06-29T06:35:47Z)
• Correcting Visual Blur Induced by Attention Distraction to Reduce Hallucinations: Algorithm and Theory [24.777344301465064]
  hallucinations are strongly associated with a human-like attention distraction phenomenon.<n>We propose an Attention-Focused Approach for Improved Image Perception, which corrects attention distraction via cross-head attention enrichment.
  arXiv  Detail & Related papers  (2026-05-23T14:36:06Z)
• Visually-Guided Policy Optimization for Multimodal Reasoning [60.035908460318126]
  We propose Visually-Guided Policy Optimization (VGPO) to reinforce visual focus during policy optimization.<n> VGPO initially introduces a Visual Attention Compensation mechanism that leverages visual similarity to localize and amplify visual cues.<n>We show that VGPO achieves better visual activation and superior performance in mathematical multimodal reasoning and visual-dependent tasks.
  arXiv  Detail & Related papers  (2026-04-10T14:22:38Z)
• Focus Matters: Phase-Aware Suppression for Hallucination in Vision-Language Models [8.304027910542446]
  We investigate the internal attention dynamics of vision encoders in Large Vision-Language Models (LVLMs)<n>Our analysis reveals that hallucination behavior is particularly sensitive to tokens receiving low attention during the focus phase.<n>We propose a lightweight inference-time intervention that selectively suppresses such tokens during the focus phase.
  arXiv  Detail & Related papers  (2026-04-04T02:46:58Z)
• Attention at Rest Stays at Rest: Breaking Visual Inertia for Cognitive Hallucination Mitigation [50.51650162235191]
  We propose a training-free Inertia-aware Visual Excitation (IVE) method that breaks this inertial pattern by modeling cognitive inference as the dynamic responsiveness of visual attention.<n>IVE is effective across various base MLLMs and multiple hallucination benchmarks, particularly for cognitive hallucinations.
  arXiv  Detail & Related papers  (2026-04-02T12:51:07Z)
• Segmentation-Based Attention Entropy: Detecting and Mitigating Object Hallucinations in Large Vision-Language Models [9.388076929154673]
  Large Vision-Language Models (LVLMs) achieve strong performance on many multimodal tasks, but object hallucinations severely undermine their reliability.<n>Most existing studies focus on the text modality, attributing hallucinations to overly strong language priors and insufficient visual grounding.<n>We propose Attention Entropy (SAE), which leverages semantic segmentation to quantify visual attention uncertainty in an object-level semantic space.
  arXiv  Detail & Related papers  (2026-03-17T14:19:22Z)
• Revealing and Enhancing Core Visual Regions: Harnessing Internal Attention Dynamics for Hallucination Mitigation in LVLMs [67.69730908817321]
  Internal Positive Attention Dynamics (PAD) in LVLMs naturally reveal semantically core visual regions under the distortions of attention sinks.<n>We propose Positive Attention Dynamics Enhancement (PADE), a training-free attention intervention that constructs a PAD map to identify semantically core visual regions.
  arXiv  Detail & Related papers  (2026-02-17T13:08:06Z)
• Cracking the Code of Hallucination in LVLMs with Vision-aware Head Divergence [69.86946427928511]
  We investigate the internal mechanisms driving hallucination in large vision-language models (LVLMs)<n>We introduce Vision-aware Head Divergence (VHD), a metric that quantifies the sensitivity of attention head outputs to visual context.<n>We propose Vision-aware Head Reinforcement (VHR), a training-free approach to mitigate hallucination by enhancing the role of vision-aware attention heads.
  arXiv  Detail & Related papers  (2024-12-18T15:29:30Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.