arXiv · cs.AI · Chih-Hsuan Yang, Anjir Ahmed Chowdhury, Cheng-Hau Yang, Weijian Zheng, Fernando Llorente, Xiaolong Ma, Xinyang Li, Eliu A. Huerta, Feng Yan, Ian T. Foster, Rajeev Thakur · 8 小时前模型评分44
研究：多智能体消息中错误答案也有用，轨迹价值应独立于答案正确性衡量
Wrong but Useful: Trajectory Value Beyond Answer Correctness in Multi-Agent Messages
导读
一项 arXiv 研究提出用“轨迹价值”衡量多智能体消息对接收方的贡献，即缓存消息存在与移除时固定整合器答对概率的变化。在 Gemma 上，73 条错误答案消息中有 34 条通过预注册复测确认，超过 11.4–20.8 的对照误确认上限；按消息移除规则比最强纯正确性规则准确率高 1.61 个百分点。gpt-oss 上证据较弱，且现有预测指标尚无法提前识别有用错误。
分类置信度
中 · 待人工复核
- 只有单一媒体来源，需要人工复核。
来源级别：媒体与个人
事件类型：研究发布
证据摘录
- “We measure this contribution as trajectory value: the change in a fixed integrator's probability of answering correctly when a cached message is present rather than removed.”
- “The finding is benchmark-dependent and the conservative criterion no longer passes without SciBench.”
- “Evidence for gpt-oss is weaker, and the tested predictors do not establish a practical way to identify useful errors in advance.”
来源：arXiv · cs.AI · arxiv.org