论文

研究:强智能体在ML工程中需要多少约束

How Much of a Harness Does a Strong Agent Need for Autonomous ML Engineering?

精选理由

论文发现复杂约束系统在MLE任务中不如简单编码基线,LLM backbone才是性能主要驱动因素。

该论文研究了自主机器学习工程(MLE)智能体在相同时间预算和前沿LLM backbone下,复杂约束系统与最小约束编码基线的性能对比。

正文

研究通过大规模系统性消融实验发现,开源的最先进约束系统在MLE基准测试中并未展现出优势。结果表明,当前MLE基准测试中,围绕强模型精心设计的约束层收益甚微。

原文 · arXiv cs.AI

How Much of a Harness Does a Strong Agent Need for Autonomous ML Engineering?

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 execution environment through read, write, and bash primitives - has received little attention in the field. In this paper we find that, under an equal time budget and the same frontier LLM backbone, open-source state-of-the-art harnesses provide no advantages over a single session of a minimal-harness coding agent baseline, pointing to the backbone as the primary driver for performance. Via a series of large-scale systematic ablation studies, we argue that the machinery layers become redundant in the coding agent setting. We conclude that the effort spent elaborating hand-crafted harnesses around strong models yields poor returns for current MLE benchmarks.

产业坐标与信源核验

本条情报归属 AITOP「论文」核心追踪栏目。在跨区域产业观察坐标中,该动态直接关联「AI 开发者工具链、Coding Agent 与工程化实践」研究主线,反映出当前技术演进与商业落地的最新风向。涉及核心实体与技术议题包括:MLE、autonomous agents、LLM、机器学习工程,读者可通过下方标签进入对应专题页查阅历史脉络与关联对比。

信源核验说明:本文原始线索采编自「arXiv cs.AI」(@Kirill Brilliantov, Alejandro Hernández-Cano, Emmanuel Abbé),经 AITOP 多维评分管线结构化提炼。截至目前,全网已有 2 家独立信源对该事件进行交叉印证与跟进报道(含 Apple ML Research 等),属于近期高关注度行业信号。

  • Apple ML Research10-01 00:00原文