MIDA Benchmark Tests Multimodal Models on Deception in Multiplayer Games
A University of Tokyo research team introduced MIDA, a benchmark that evaluates whether multimodal models can distinguish truth from deception in multiplayer social interactions. It contains 2,360 utterances from Ego4D and YouTube game footage, with video, audio, dialogue history, game rules, and verifiable true, false, or neutral labels. The reported experiments found that even strong models struggled with binary truthfulness judgments, while dialogue history was more useful than additional visual frames.
MIDA provides a controlled way to measure the social and mind-reasoning capabilities needed for multiplayer collaboration, human-computer interaction, and multi-agent systems.
新智元 2026-10-28 19:50 北京
新智元报道
人类说一句「我不是狼」，多模态大模型既能看到表情，也能听见声音。但它真的知道这句话是真是假吗？
最近，来自东京大学的研究团队提出了多模态交互式欺骗评估基准 MIDA（Multimodal Interactive Deception Assessment），系统测试 GPT、Gemini、Claude、Qwen、DeepSeek 等开源和闭源模型在多人社交互动中的欺骗判断能力。
结果显示，即使是当前能力较强的多模态大模型，也很难可靠地区分真话与谎言。它们能够理解图像、声音和文本，却未必能弄清楚一个人在当时「知道什么、相信什么、又试图让别人相信什么」。
#MIDA#University of Tokyo#GPT-4o#GPT-5#GPT-4o-mini#Gemini#Claude#Qwen2.5-VL-7B#DeepSeek#Llama#InternVL#Ego4D