通过对抗指令调优缓解大型视觉语言模型的对话幻觉
Mitigating Dialogue Hallucination for Large Vision Language Models via Adversarial Instruction Tuning
AI总结：
针对多轮对话诱发LVLMs幻觉的问题，本文构建AQG对抗对话评估基准，揭示模型偏向对话历史的成因，并提出AIT对抗指令调优，在保持性能的同时显著降低对话幻觉。
摘要：
缓解大型视觉语言模型（LVLMs）的幻觉对于提升其作为通用助手的可靠性至关重要。本文表明，先前的用户—系统对话会显著加剧LVLMs的此类幻觉。为精确衡量这一问题，我们首先提出一个评估基准：利用我们新颖的Adversarial Question Generator（AQG，对抗性问题生成器），在流行的多模态基准数据集前预置幻觉对话，从而扩展这些数据集；AQG通过对LVLMs实施对抗攻击，可自动生成与图像相关但具有对抗性的对话。在我们的基准上，最先进LVLMs在VQA和Captioning任务中的零样本性能均显著下降。随后，我们进一步揭示这种幻觉主要源于模型对先前对话而非视觉内容的预测偏差。为减少该偏差，我们提出Adversarial Instruction Tuning（AIT，对抗指令调优），针对幻觉对话对LVLMs进行鲁棒微调。大量实验表明，我们提出的方法成功减少了对话幻觉，同时保持了原有性能。
英文摘要：
Mitigating hallucinations of Large Vision Language Models,(LVLMs) is crucial to enhance their reliability for general-purpose assistants. This paper shows that such hallucinations of LVLMs can be significantly exacerbated by preceding user-system dialogues. To precisely measure this, we first present an evaluation benchmark by extending popular multi-modal benchmark datasets with prepended hallucinatory dialogues powered by our novel Adversarial Question Generator (AQG), which can automatically generate image-related yet adversarial dialogues by adopting adversarial attacks on LVLMs. On our benchmark, the zero-shot performance of state-of-the-art LVLMs drops significantly for both the VQA and Captioning tasks. Next, we further reveal this hallucination is mainly due to the prediction bias toward preceding dialogues rather than visual content. To reduce this bias, we propose Adversarial Instruction Tuning (AIT) that robustly fine-tunes LVLMs against hallucinatory dialogues. Extensive experiments show our proposed approach successfully reduces dialogue hallucination while maintaining performance.
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。
