通过多智能体辩论解释和缓解 MLLMs 中的幻觉
Interpreting and Mitigating Hallucination in MLLMs through Multi-agent Debate
AI总结：
针对 MLLMs 幻觉问题，提出基于自我反思的慢思考机制与多智能体辩论方法，不仅能缓解并解释幻觉成因，还区分了创造力与幻觉，在多基准上展现出泛化性能。
摘要：
MLLMs 经常生成与视觉内容不一致的输出，这一挑战被称为幻觉。以往的方法侧重于判断生成的输出是否为幻觉，而没有识别导致幻觉的图像区域，也没有解释为何会产生此类幻觉。在本文中，我们认为 MLLMs 中的幻觉部分归因于这些模型缺乏慢思考和发散性思考。为了解决这个问题，我们提出采用自我反思机制来促进慢思考。此外，我们将消除幻觉视为一项复杂的推理任务，并提出一种多智能体辩论方法来鼓励发散性思考。因此，我们的方法不仅能缓解幻觉，还能解释幻觉产生的原因并详述幻觉的具体细节。此外，我们提出在 MLLMs 语境下区分创造力与幻觉，并说明了如何评估 MLLMs 的创造力能力。在多个基准上的大量实验表明，我们的方法在多个 MLLMs 上展现出泛化的幻觉缓解性能。
英文摘要：
MLLMs often generate outputs that are inconsistent with the visual content, a challenge known as hallucination. Previous methods focus on determining whether a generated output is hallucinated, without identifying which image region leads to the hallucination or interpreting why such hallucinations occur. In this paper, we argue that hallucination in MLLMs is partially due to a lack of slow-thinking and divergent-thinking in these models. To address this, we propose adopting a self-reflection scheme to promote slow-thinking. Furthermore, we consider eliminating hallucination as a complex reasoning task and propose a multi-agent debate approach to encourage divergent-thinking. Consequently, our approach can not only mitigate hallucinations but also interpret why they occur and detail the specifics of hallucination. In addition, we propose to distinguish creativity from hallucination in the context of MLLMs, and illustrate how to evaluate MLLMs' creativity capability. Extensive experiments on various benchmarks demonstrate that our approach exhibits generalized hallucinations-mitigating performance across several MLLMs.
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。
