边际价值估计：高效深度研究智能体的剪枝策略研究
原标题：Not Worth Another Token: Marginal Value Estimation for Efficient Deep Research Agents
AI 摘要
Hugging Face 每日论文发布了一篇关于深度研究智能体效率优化的研究论文。论文提出在长周期研究智能体的不同流水线阶段应用剪枝策略，以减少 token 使用和延迟，并发现早期剪枝能带来最大的效率提升。实验表明，轻量级启发式方法可减少高达 73% 的 token 使用且质量损失很小，但没有单一方法在所有指标上占优。该研究为设计高效的长周期智能体系统提供了实用指导。
正文节选
Not Worth Another Token: Marginal Value Estimation for Efficient Deep Research Agents Abstract Pruning strategies applied at different pipeline stages reduce token usage and latency in long-horizon research agents, with early pruning yielding the greatest efficiency gains. Long-horizon research agents solve open-ended tasks through iterative retrieval, aggregation, and synthesis, but context grows rapidly while the marginal value of additional evidence often declines. This leads to unnecessary t