Princeton 与 Google 论文提出 JevAgentHarness，用 4B 决策模型分担 LLM 智能体例行决策
X：Rohan Paul (@rohanpaul_ai)AIHOT score 58/100
Summary
Princeton 与 Google 等机构的论文提出 JevAgentHarness，用小型决策模型回答智能体中的多选例行决策，仅在置信度低于阈值时交给 LLM。
Original title
New Princeton Univ + Google paper finds that the decision model most accurate on isolated questions solved only 69.5% of agent tasks, wh...