arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2610.37201cs.AI
收藏

你的通用技能优化器真的能激励智能体任务吗?通过二分归因重新思考技能优化

Does Your Generalist Skill Optimizer Really Incentivize Agentic Tasks? Rethinking Skill Optimization via Bipartite Attribution

Shanghai AI Laboratory(上海人工智能实验室);Fudan University(复旦大学);Shanghai Jiao Tong University(上海交通大学);University of Science and Technology of China(中国科学技术大学);Tsinghua University(清华大学)

Jiawei Zhou, Bo Peng, Yuxuan Huang, Ming Zhang, Youbang Sun, Xiang Wang, Chaochao Lu

AI总结:

本研究针对技能优化器在智能体任务中的激励效果问题,提出BiPA方法,通过二分归因分析发现其开发集增益在测试集消失,进而给出四条优化原则。

摘要:

技能优化器通过重写智能体的技能(即其上下文内的自然语言过程)来使智能体适应长期任务,例如修复终端环境,并保留能提升小型开发集上成功率的编辑。在这类任务中,成功率存在噪声,因此小型开发集上的增益无法证明某一编辑有帮助,且成功得分无法体现该编辑的代价。我们提出BiPA,它会在与种子完全相同的候选样本上重放优化器的接受规则,以衡量仅选择机制会报告什么,并将每个回合记录为轮次与智能体读取文本的二分图,从而追踪额外输入令牌与背后行为的关联。在三个优化器、三个智能体基准和两个模型上,学习到的技能成功率与手动编写的种子相当,平均高出0.1个点。它们在测试集上的开发集增益消失:在至多30个任务中保留多个候选样本中的最佳者,即使没有任何编辑改变,也会报告如此大的增益。它们读取的输入令牌增加了约16%:优化器几乎总是添加规则,其中一半用于重新检查工作,部分用于一次调用执行一个操作,智能体将这些规则作为额外轮次遵循,每次都会重新发送整个对话。由此得出四条原则:仅当编辑的增益超过仅选择机制报告的数值时才接受该编辑;按编辑添加的轮次而非长度计费;允许编辑器删除和合并规则;向其展示早期规则的代价。代码可在该https URL获取。

英文摘要:

Skill optimizers adapt an agent to long-horizon tasks, such as repairing a terminal environment, by rewriting its skill, a natural-language procedure in its context, and keeping the edits that raise success on a small dev set. On such tasks success is noisy, so a gain on a small dev set does not show that an edit helps, and a success score does not show what the edit costs. We introduce BiPA, which replays an optimizer's acceptance rule on candidates identical to its seed to measure what selection alone reports, and records each episode as a bipartite graph of turns and the text they read to trace extra input tokens to the behavior behind them. Across three optimizers, three agentic benchmarks and two models, the learned skills succeed about as often as their hand-written seed, 0.1 points higher on average. Their dev gains vanish on test: keeping the best of several candidates on at most 30 tasks reports gains this large even when none changed. They read about 16% more input tokens: the optimizers almost always add rules, half to re-check work and some to act one call at a time, and the agent follows them as extra turns, each re-sending the whole conversation. Four principles follow: accept an edit only if its gain exceeds what selection alone reports; charge it for the turns it adds, not for its length; let the editor delete and merge rules; show it what earlier rules cost. Code is available at https://github.com/valleysprings/BiPA.git.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑