[2605.17373] FML-bench: A Controlled Study of AI Research Agent Strategies from the Perspective of Search Dynamics

-->

Computer Science > Machine Learning

arXiv:2605.17373 (cs)

[Submitted on 17 May 2026]

Title:FML-bench: A Controlled Study of AI Research Agent Strategies from the Perspective of Search Dynamics

Authors:Qiran Zou, Hou Hei Lam, Wenhao Zhao, Tingting Chen, Yiming Tang, Samson Yu, Yingtao Zhu, Srinivas Anumasa, Zufeng Zhang, Tianyi Zhang, Chang Liu, Zhengyao Jiang, Anirudh Goyal, Dianbo Liu<br>View a PDF of the paper titled FML-bench: A Controlled Study of AI Research Agent Strategies from the Perspective of Search Dynamics, by Qiran Zou and 13 other authors

View PDF<br>HTML (experimental)

Abstract:AI research agents accelerate ML research by automating hypothesis generation, experimentation, and empirical refinement. Existing agent strategies range from greedy hill-climbing to tree search and evolutionary optimization, yet which strategy choices drive performance remains unclear. Answering this question requires a benchmark that separates agent strategy (e.g., search topology) from execution infrastructure (e.g., code editor), so that performance differences are attributable to strategy rather than infrastructure, and that provides process-level metrics beyond final scores to analyze exploration behaviors. Existing benchmarks offer limited support. We propose FML-Bench, a benchmark of 18 fundamental ML research tasks across 10 domains that separates agent strategy from execution infrastructure and defines 12 process-level behavioral metrics. Evaluating six representative agents, we find that: (1) strategy complexity alone does not guarantee strong performance: a simple greedy hill-climber nearly matches the best-performing tree-search agent, both well above the remaining agents; (2) our analysis suggests this pattern relates to improvement opportunity structure: greedy search tends to be more effective when opportunities are dense, while tree-search and evolutionary strategies tend to be more effective when opportunities are sparse; an adaptive agent built on this insight switches to broader exploration upon detecting improvement stagnation and outperforms the other six agents, lending initial support to this observation; and (3) process-level analysis reveals that early convergence and directionally focused exploration are significantly associated with final performance, while solution diversity and compute cost are not. Our benchmark is available at: this https URL.

Comments:<br>Our benchmark is available at: this https URL

Subjects:

Machine Learning (cs.LG); Artificial Intelligence (cs.AI)

Cite as:<br>arXiv:2605.17373 [cs.LG]

(or<br>arXiv:2605.17373v1 [cs.LG] for this version)

https://doi.org/10.48550/arXiv.2605.17373

Focus to learn more

arXiv-issued DOI via DataCite (pending registration)

Submission history<br>From: Qiran Zou [view email]<br>[v1]<br>Sun, 17 May 2026 10:30:38 UTC (1,413 KB)

Full-text links:<br>Access Paper:

View a PDF of the paper titled FML-bench: A Controlled Study of AI Research Agent Strategies from the Perspective of Search Dynamics, by Qiran Zou and 13 other authors<br>View PDF<br>HTML (experimental)<br>TeX Source

view license

Current browse context:

cs.LG

next >

new<br>recent<br>| 2026-05

Change to browse by:

cs<br>cs.AI

References & Citations

NASA ADS<br>Google Scholar

Semantic Scholar

export BibTeX citation<br>Loading...

BibTeX formatted citation

&times;

loading...

Data provided by:

Bookmark

Bibliographic Tools

Bibliographic and Citation Tools

Bibliographic Explorer Toggle

Bibliographic Explorer (What is the Explorer?)

Connected Papers Toggle

Connected Papers (What is Connected Papers?)

Litmaps Toggle

Litmaps (What is Litmaps?)

scite.ai Toggle

scite Smart Citations (What are Smart Citations?)

Code, Data, Media

Code, Data and Media Associated with this Article

alphaXiv Toggle

alphaXiv (What is alphaXiv?)

Links to Code Toggle

CatalyzeX Code Finder for Papers (What is CatalyzeX?)

DagsHub Toggle

DagsHub (What is DagsHub?)

GotitPub Toggle

Gotit.pub (What is GotitPub?)

Huggingface Toggle

Hugging Face (What is Huggingface?)

ScienceCast Toggle

ScienceCast (What is ScienceCast?)

Demos

Demos

Replicate Toggle

Replicate (What is Replicate?)

Spaces Toggle

Hugging Face Spaces (What is Spaces?)

Spaces Toggle

TXYZ.AI (What is TXYZ.AI?)

Related Papers

Recommenders and Search Tools

Link to Influence Flower

Influence Flower (What are Influence Flowers?)

Core recommender toggle

CORE Recommender (What is CORE?)

IArxiv recommender toggle

IArxiv Recommender<br>(What is IArxiv?)

Author

Venue

Institution

Topic

About arXivLabs

arXivLabs: experimental projects with community collaborators

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv...