HalluShift++: Bridging Language and Vision through Internal Representation Shifts for Hierarchical Hallucinations in MLLMs

TL;DR

HalluShift++ introduces a novel internal layer analysis method to detect hallucinations in multimodal large language models, improving factual consistency assessment without relying solely on external evaluators.

Contribution

The paper presents a new approach that analyzes internal layer dynamics of MLLMs to detect hallucinations, extending techniques from text-based LLMs to multimodal models.

Findings

Effective detection of hallucinations through internal layer irregularities.

Broadens hallucination assessment from text-only to multimodal models.

Codebase available for implementation and further research.

Abstract

Multimodal Large Language Models (MLLMs) have demonstrated remarkable capabilities in vision-language understanding tasks. While these models often produce linguistically coherent output, they often suffer from hallucinations, generating descriptions that are factually inconsistent with the visual content, potentially leading to adverse consequences. Therefore, the assessment of hallucinations in MLLM has become increasingly crucial in the model development process. Contemporary methodologies predominantly depend on external LLM evaluators, which are themselves susceptible to hallucinations and may present challenges in terms of domain adaptation. In this study, we propose the hypothesis that hallucination manifests as measurable irregularities within the internal layer dynamics of MLLMs, not merely due to distributional shifts but also in the context of layer-wise analysis of specific assumptions. By incorporating such modifications, \textsc{\textsc{HalluShift++}} broadens the efficacy of hallucination detection from text-based large language models (LLMs) to encompass multimodal scenarios. Our codebase is available at https://github.com/C0mRD/HalluShift_Plus.