ReactBench：通过系统评估的多模态幻觉因果驱动基准
ReactBench: A Cause-Driven Benchmark for Multimodal Hallucination via Systematic Evaluation
AI总结：
提出ReactBench基准，通过对抗性图像和诱导幻觉的查询，系统评估多模态大模型在关系擦除、反事实属性、变化追踪和密集计数等任务中的因果幻觉。
摘要：
尽管多模态大语言模型（MLLMs）在视觉-语言理解方面取得了快速进展，但它们仍然容易产生多模态幻觉，即生成与视觉输入不一致的响应。现有基准主要侧重于检测幻觉结果，而非评估这些失败的潜在原因。此外，许多基准依赖于简单的场景和有限的评估格式，不再能挑战最先进的模型。为了解决这些局限性，我们引入了ReactBench，一个因果驱动的幻觉基准，具有多个任务和考试式评估格式。通过生成对抗性图像和诱导幻觉的查询，ReactBench引入了四个目标任务：关系擦除、反事实属性、变化追踪和密集计数。这些任务系统地暴露了共现偏差、语言先验、跨图像比较感知缺陷和细粒度感知瓶颈。除了基于标准准确率的评估外，我们利用思维链推理来识别每个任务中幻觉的细粒度子原因。大量评估表明，当前的MLLMs仍然容易受到特定因果幻觉触发因素的影响，这证明了ReactBench作为诊断和提高多模态模型鲁棒性的系统化和可解释测试平台的价值。项目页面见https://reactbench.github.io/。
英文摘要：
While multimodal large language models (MLLMs) have achieved rapid progress in vision-language understanding, they remain prone to multimodal hallucinations, producing responses that are inconsistent with the visual input. Existing benchmarks predominantly focus on detecting hallucination outcomes rather than evaluating the underlying causes of these failures. Moreover, many benchmarks rely on simplistic scenarios and limited evaluation formats that no longer challenge state-of-the-art models. To address these limitations, we introduce ReactBench, a cause-driven hallucination benchmark featuring multiple tasks and an exam-style evaluation format. By generating adversarial images and hallucination-inducing queries, ReactBench introduces four targeted tasks: Relational Erasure, Counterfactual Attribute, Alteration Tracing, and Dense Counting. These tasks systematically expose co-occurrence bias, language priors, cross-image comparative perception deficiencies, and fine-grained perceptual bottlenecks. Beyond standard accuracy-based evaluation, we leverage Chain-of-Thought reasoning to identify fine-grained sub-causes of hallucination within each task. Extensive evaluations reveal that current MLLMs remain notably vulnerable to cause-specific hallucination triggers, demonstrating the value of ReactBench as a systematic and interpretable testbed for diagnosing and improving multimodal model robustness. The project page is available at https://reactbench.github.io/.
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。
