强智能体做自主ML工程需要多少框架？
原标题：How Much of a Harness Does a Strong Agent Need for Autonomous ML Engineering?
AI 摘要
Apple Machine Learning Research 发表论文，研究自主机器学习工程（MLE）智能体是否需要复杂的工程框架（harness）。研究发现，在相同时间预算和相同前沿 LLM 基座下，开源最先进的复杂框架相比仅使用读写和 bash 原语的最小化编码智能体基线并无优势，性能主要由基座模型决定。通过大规模系统性消融实验，作者认为在编码智能体场景下，多智能体编排、检索子智能体等机制层是冗余的，围绕强模型手工构建复杂框架对当前 MLE 基准收益甚微。
正文节选
Recent autonomous machine learning engineering (MLE) agents have made significant progress on public leaderboards. Often motivated by progress stagnation over long-horizon cycles and limited Large Language Model (LLM) primitives, modern MLE agents are deployed on top of increasingly elaborate machinery: multi-agent orchestrators, dedicated retrieval subagents, and more. While such harnesses expand, the use of more primitive but improved coding agents—where LLMs have direct access to the executio