UCLA 与东北大学研究：AI 智能体约 18% 输出 token 属可移除浪费
UCLA研究：AI智能体约18%输出token是浪费
UCLA 与东北大学论文发现，AI 智能体约 18% 的输出 token 是可移除的浪费，剪掉后任务成功率几乎不受影响。研究覆盖 6 个前沿模型和 4 个 benchmark，方法是将智能体调用经代理转发，由 LLM 判定可删除片段后重跑验证。
Rohan Paul· @rohanpaul_ai · X·2026-11-09 20:14· 2 小时前AI 评分46
AI 导读
UCLA与东北大学论文发现，AI智能体约18%的输出token属可移除浪费，剪掉后任务成功率几乎不受影响。研究覆盖6个前沿模型和4个benchmark，在Kimi-K3的推理中剪除重复与填充内容后，SWE-bench Verified上减少41%输出token，成功率89.5%对89.0%。作者建议按任务与harness定位浪费来源，而非单纯限制token。
正文 · 原文
New UCLA+ Northeastern University paper finds that about 18% of AI agent output tokens are removable waste, and cutting it barely hurts task success.
Agent bills are usually measured as total tokens. That number can't tell you which tokens did useful work and which just repeated a policy, refetched a file, or went down a dead end.
They route agent calls through a proxy and have an LLM judge mark spans that can go without losing useful work. Then they delete those spans and rerun the agent to check nothing breaks.
Across 6 frontier models and 4 benchmarks, about 18% of output tokens were waste. Pruning repetition and filler from Kimi-K3's reasoning removed 41% of it on SWE-bench Verified, and success stayed at 89.5% versus 89.0%.
To cut agent costs, find where your task and harness create waste instead of just capping tokens.
来源：Rohan Paul · x.com
来源：RadarAI · aihot.news