MLLMs Hallucinate when Information Distribution Drifts in Synergy Heads
Multimodal Large Language Models (MLLMs) often struggle with hallucinations, thus hindering their reliable practical applications. Existing attention-based mitigation methods mainly rely on indirect signals (e.g., attention weights) that fail to accurately reflect the actual information shift underlying hallucination generation. In this paper, we propose HEAL, Head-lEvel information disentAnglement and caLibration for identifying and mitigating hallucinations. HEAL first employs causal noise intervention on multi-head outputs to filter out causally redundant heads. Subsequently, it disentangles information distribution within the remaining heads via the counterfactual Difference-in-Differences, categorizing heads into four types. Through analysis, we observe: hallucinations happen when information distribution drifts away from a healthy equilibrium in synergy heads, not strongly correlated with the quantity or strength of modality-specific heads. Motivated by this insight, HEAL injects dynamic information calibration factors into the value vectors of synergy heads, and actively regulates visual-language dependencies, steering the output distribution towards factual evidence. Extensive experiments demonstrate that HEAL effectively reduces hallucinations across multiple MLLMs, offering a simple and interpretable pathway to enhance model trustworthiness.
Code (0)
등록된 구현이 없습니다.
Similar Papers 제목 키워드 기반
Data-augmented phrase-level alignment for mitigating object hallucination
Despite their significant advancements, Multimodal Large Language Models (MLLMs) often generate factually inaccurate information, referred to as hallucination. In this work, we address object hallucinations in MLLMs, whe…
Data AugmentationHallucinationImage DescriptionObject+3Drawing the Line: Enhancing Trustworthiness of MLLMs Through the Power of Refusal
Multimodal large language models (MLLMs) excel at multimodal perception and understanding, yet their tendency to generate hallucinated or inaccurate responses undermines their trustworthiness. Existing methods have large…
FINER: MLLMs Hallucinate under Fine-grained Negative Queries
Multimodal large language models (MLLMs) struggle with hallucinations, particularly with fine-grained queries, a challenge underrepresented by existing benchmarks that focus on coarse image-related questions. We introduc…
Mitigating Hallucinations in Multimodal LLMs via Object-aware Preference Optimization
Multimodal Large Language Models (MLLMs) emerge as a unified interface to address a multitude of tasks, ranging from NLP to computer vision. Despite showcasing state-of-the-art results in many benchmarks, a long-standing…
Image CaptioningConVis: Contrastive Decoding with Hallucination Visualization for Mitigating Hallucinations in Multimodal Large Language Models
Hallucinations in Multimodal Large Language Models (MLLMs) where generated responses fail to accurately reflect the given image pose a significant challenge to their reliability. To address this, we introduce ConVis, a n…
Hallucination