Not Worth Another Token: Marginal Value Estimation for Efficient Deep Research Agents
- 类型:arxiv
- 标识:2608.08389
- 链接:https://arxiv.org/abs/2608.08389
- 主分类:agent
- 形态:method
- 被引:0
- 被引来源:Semantic Scholar
- S2被引:0
- 影响力被引:0
- TLDR:The results show that pruning effectiveness depends more on where pruning is applied than on the specific scoring rule: early pruning yields the largest end-to-end savings, while later pruning mainly refines the final synthesis context.
- 副分类:rag
- 待LLM分类:否
- 标题中文:不值得再花一个 token:面向高效深度研究 Agent 的边际价值估计
- TLDR中文:结果表明,剪枝效果更取决于剪枝应用的位置,而非具体的评分规则:早期剪枝带来最大的端到端节省,后期剪枝主要用于细化最终的合成上下文。
- 来源文件:
- /inbox/tom/_candidates/2026-08-12-agent-rag-longcontext-candidates.json
- [S2 enrich]