← 研踪 PaperTrailarXiv:2611.25357 · v1
Pro$^2$ Agent: Scientific Problem Probing for Autonomous AI R&D
首次提交 2026-11-19 09:52
版本更新 2026-11-19 09:52
雷达观测 2026-11-20 10:50
摘要
Large language model (LLM) agents can now autonomously conduct AI research by writing code and running experiments. They typically do so by iteratively optimizing a codebase, while ignoring the problem structure behind the task. This formulation is effective on simpler tasks, but on difficult problems it often gets stuck in the wrong solution region because the agent never explicitly represents its understanding of the problem. We introduce Pro Agent, which shifts the primary state of autonomous AI research from a codebase state to a Problem State: an explicit structure of what problems prevent success and which solutions are likely to address them. Pro Agent initializes the Problem State through comprehensive analysis of available resources and constraints. It then performs targeted research to relate the identified problems to candidate solutions. Through controlled hypothesis testing, the agent updates the importance of a suspected problem or the benefit brought by a candidate solution. On the complete MLE-Bench, Pro Agent matches the best reported medal rate (70.7%) under the standard 24-hour budget and reaches 76.0% with a 48-hour budget, the highest medal rate reported on the complete benchmark. It also achieves competitive performance on PostTrainBench. In addition, it produces research notes for each task that make the research process transparent and reusable by human researchers.
观测到的版本
相邻记录可能跳过中间版本;首次发现 v2 及以上也会标为版本更新。
v1 · 2026-11-19 09:52 · 最新观测版本
Pro$^2$ Agent: Scientific Problem Probing for Autonomous AI R&D
Large language model (LLM) agents can now autonomously conduct AI research by writing code and running experiments. They typically do so by iteratively optimizing a codebase, while ignoring the problem structure behind the task. This formulation is effective on simpler tasks, but on difficult problems it often gets stuck in the wrong solution region because the agent never explicitly represents its understanding of the problem. We introduce Pro Agent, which shifts the primary state of autonomous AI research from a codebase state to a Problem State: an explicit structure of what problems prevent success and which solutions are likely to address them. Pro Agent initializes the Problem State through comprehensive analysis of available resources and constraints. It then performs targeted research to relate the identified problems to candidate solutions. Through controlled hypothesis testing, the agent updates the importance of a suspected problem or the benefit brought by a candidate solution. On the complete MLE-Bench, Pro Agent matches the best reported medal rate (70.7%) under the standard 24-hour budget and reaches 76.0% with a 48-hour budget, the highest medal rate reported on the complete benchmark. It also achieves competitive performance on PostTrainBench. In addition, it produces research notes for each task that make the research process transparent and reusable by human researchers.
查看此版本解读 →查看此版本原文 ↗