mlsanityData to 14 Oct 2026

AI Research Agents for Machine Learning: Search, Exploration, and Generalization in MLE-bench

arXiv 2507.02554 ↗First posted 3 Jul 2025cs.AI cs.LGcited by 71

On the map: LLM agents and tool use → Web, search, and research agents

Abstract (arXiv)

AI research agents are demonstrating great potential to accelerate scientific progress by automating the design, implementation, and training of machine learning models. We focus on methods for improving agents' performance on MLE-bench, a challenging benchmark where agents compete in Kaggle competitions to solve real-world machine learning problems. We formalize AI research agents as search policies that navigate a space of candidate solutions, iteratively modifying them using operators. By designing and systematically varying different operator sets and search policies (Greedy, MCTS, Evolutionary), we show that their interplay is critical for achieving high performance. Our best pairing of search strategy and operator set achieves a state-of-the-art result on MLE-bench lite, increasing the success rate of achieving a Kaggle medal from 39.6% to 47.7%. Our investigation underscores the importance of jointly considering the search strategy, operator design, and evaluation methodology in advancing automated machine learning.

Key facts. Cited by 71 papers in the map (data through 14 Oct 2026). Placed in Web, search, and research agents, one of 16 directions in LLM agents and tool use. Builds on 31 of its 56 references that are in the map. First posted on arXiv 3 Jul 2025.

Builds on

31 of 56 references.

Closest ancestors · 6

Same direction, newest first

  1. 2025The AI Scientist-v2: Workshop-Level Automated Scientific Discovery via Agentic Tree Search10 Apr
  2. 2025MLGym: A New Framework and Benchmark for Advancing AI Research Agents20 Feb
  3. 2024RE-Bench: Evaluating frontier AI R&D capabilities of language model agents against human experts22 Nov
  4. 2024MLE-bench: Evaluating Machine Learning Agents on Machine Learning Engineering9 Oct
  5. 2024Can LLMs Generate Novel Research Ideas? A Large-Scale Human Study with 100+ NLP Researchers6 Sep
  6. 2023MLAgentBench: Evaluating Language Agents on Machine Learning Experimentation5 Oct

Other references · 25

  1. 2022Chain-of-Thought Prompting Elicits Reasoning in Large Language Modelscited by 12,459
  2. 2022ReAct: Synergizing Reasoning and Acting in Language Modelscited by 8,695
  3. 2025DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learningcited by 6,330
  4. 2022Self-Consistency Improves Chain of Thought Reasoning in Language Modelscited by 5,708
  5. 2023Tree of Thoughts: Deliberate Problem Solving with Large Language Modelscited by 3,211
  6. 2024Scaling LLM Test-Time Compute Optimally can be More Effective than Scaling Model Parameterscited by 1,986
  7. 2018DARTS: Differentiable Architecture Searchcited by 1,798
  8. 2024SWE-agent: Agent-Computer Interfaces Enable Automated Software Engineeringcited by 1,737
  9. 2016Neural Architecture Search with Reinforcement Learningcited by 1,499
  10. 2018Regularized Evolution for Image Classifier Architecture Searchcited by 909
  11. 2023Reasoning with Language Model is Planning with World Modelcited by 861
  12. 2016Hyperband: A Novel Bandit-Based Approach to Hyperparameter Optimizationcited by 699
13 more
  1. 2023Tool Learning with Foundation Modelscited by 302
  2. 2025AIDE: AI-Driven Exploration in the Space of Codecited by 226
  3. 2024Large Language Model-Based Agents for Software Engineering: A Surveycited by 178
  4. 2020Auto-Sklearn 2.0: Hands-free AutoML via Meta-Learningcited by 174
  5. 2024RLEF: Grounding Code LLMs in Execution Feedback with Reinforcement Learningcited by 151
  6. 2024SWE-Search: Enhancing Software Agents with Monte Carlo Tree Search and Iterative Refinementcited by 132
  7. 2024MegaScale: Scaling Large Language Model Training to More Than 10,000 GPUscited by 112
  8. 2024OMNI-EPIC: Open-endedness via Models of human Notions of Interestingness with Environments Programmed in Codecited by 74
  9. 2023SayCanPay: Heuristic Planning with Large Language Models using Learnable Domain Knowledgecited by 49
  10. 2024Discovering Preference Optimization Algorithms with and for Large Language Modelscited by 42
  11. 2017Simple And Efficient Architecture Search for Convolutional Neural Networkscited by 40
  12. 2024Revisiting Reliability in Large-Scale Machine Learning Research Clusterscited by 28
  13. 2024REvolve: Reward Evolution with Large Language Models using Human Feedbackcited by 23

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 12; 2026: 59.

71 papers.

Sort:

By area:

  1. 2025Quagmires in SFT-RL Post-Training: When High SFT Scores Mislead and What to Use Insteadcited by 28
  2. 2026Towards Execution-Grounded Automated AI Researchcited by 28
  3. 2026MLEvolve: A Self-Evolving Framework for Automated Machine Learning Algorithm Discoverycited by 28
  4. 2025RECODE-H: A Benchmark for Research Code Development with Interactive Human Feedbackcited by 25
  5. 2026Toward Generalist Autonomous Research via Hypothesis-Tree Refinementcited by 22
  6. 2025AutoMLGen: Navigating Fine-Grained Optimization for Coding Agentscited by 18
  7. 2026Toward Autonomous Long-Horizon Engineering for ML Researchcited by 17
  8. 2026AIBuildAI: An AI Agent for Automatically Building AI Modelscited by 13
  9. 2026AIRA_2: Overcoming Bottlenecks in AI Research Agentscited by 12
  10. 2026KAPSO: A Knowledge-grounded framework for Autonomous Program Synthesis and Optimizationcited by 11
  11. 2026Synthetic Sandbox for Training Machine Learning Engineering Agentscited by 11
  12. 2026Recursive self-improvement of AI research agentscited by 11
38 more
  1. 2025Bootstrapping Task Spaces for Self-Improvementcited by 10
  2. 2026Can We Predict Before Executing Machine Learning Agents?cited by 10
  3. 2026Harnessing the Collective Intelligence of AI Agents in the Wild for New Discoveriescited by 10
  4. 2026AI Research Preference Modelscited by 10
  5. 2026Simple Baselines are Competitive with Code Evolutioncited by 9
  6. 2026Training AI Scientists to Replicate Researchcited by 9
  7. 2026Understanding the Challenges in Iterative Generative Optimization with LLMscited by 8
  8. 2026FML-bench: A Controlled Study of AI Research Agent Strategies from the Perspective of Search Dynamicscited by 8
  9. 2026Frontis-MA1: Training an AI4AI Model towards Recursive Self-Improvement in Machine Learning Engineeringcited by 8
  10. 2025End-to-end PDDL Planning with Hardcoded and Dynamic Agentscited by 5
  11. 2025KernelEvolve: Scaling Agentic Kernel Coding for Heterogeneous AI Accelerators at Metacited by 5
  12. 2026Agentic Discovery of Neural Architectures: AIRA-Compose and AIRA-Designcited by 5
  13. 2026AceGRPO: Adaptive Curriculum Enhanced Group Relative Policy Optimization for Autonomous Machine Learning Engineeringcited by 4
  14. 2026Reasoning as Gradient: Scaling MLE Agents Beyond Tree Searchcited by 4
  15. 2026GEAR: Genetic AutoResearch for Agentic Code Evolutioncited by 4
  16. 2026ScienceFlow: A long-horizon agent for ML research, scientific discovery and beyondcited by 4
  17. 2025Illuminating LLM Coding Agents: Visual Analytics for Deeper Understanding and Enhancementcited by 3
  18. 2025ML-Tool-Bench: Tool-Augmented Planning for ML Taskscited by 3
  19. 2025ReX-MLE: The Autonomous Agent Benchmark for Medical Imaging Challengescited by 3
  20. 2026AIBuildAI-2: A Knowledge-Enhanced Agent for Automatically Building AI Modelscited by 3
  21. 2026Why Solve It Twice? Hierarchical Accumulation of Skills for Transfer-Efficient ML Engineeringcited by 3
  22. 2026RecHarness: A Bandit-Routed Agentic Harness for Self-Evolving Recommender Systemscited by 3
  23. 2025SelfAI: A self-directed framework for long-horizon scientific discoverycited by 2
  24. 2026Heuresis: Search Strategies for Autonomous AI Research Agents Across Quality, Diversity and Noveltycited by 2
  25. 2026Matryoshka Agent: Unfolding Sub-Agents for Long-Horizon Machine Learning Engineeringcited by 2
  26. 2026Beyond Solution-Centric Search: Adaptive Inquiry and Knowledge Revision for Autonomous ML Engineeringcited by 2
  27. 20262AM: Grounding Agent-Side Memory as Guidance for Steerable Action Models in Long-Horizon Manipulationcited by 2
  28. 2025What Makes AI Research Replicable? Executable Knowledge Graphs as Scientific Knowledge Representationscited by 1
  29. 2026SciDER: Scientific Data-centric End-to-end Researchercited by 1
  30. 2026stratum: A System Infrastructure for Massive Agent-Centric ML Workloadscited by 1
  31. 2026DrugSAGE:Self-evolving Agent Experience for Efficient State-of-the-Art Drug Discoverycited by 1
  32. 2026Be Fair! Can Machine Learning Engineering Agents Adhere to Fairness Constraints?cited by 1
  33. 2026Muse Spark Safety & Preparedness Reportcited by 1
  34. 2026Learning the ARTS of Search for Automated Discoverycited by 1
  35. 2026Discovering Crystal Structure Prediction Algorithms with an AI Co-Scientistcited by 1
  36. 2026IDEAgent: Agentic Quality-Diversity Search for Research Idea Generationcited by 1
  37. 2026Idea Search: Guiding Tree Search with Ideas to Explore Diverse Scientific Methodscited by 1
  38. 2026The Greatness of Science Cannot Be Planned: Agentic Auto-Research is Fuzz Testingcited by 1

50 most cited of 71 All 71 (JSON)

Similar, not cited

Close in wording, no citation either way

  1. 2025MLR-Bench: Evaluating AI Agents on Open-Ended Machine Learning Research
  2. 2026ML-AutoResearch: Training Machine Learning Research Agents with Automatically Generated Environments
  3. 2025CoMind: Towards Community-Driven Agents for Machine Learning Engineering
  4. 2025R&D-Agent: An LLM-Agent Framework Towards Autonomous Data Science
  5. 2025FML-bench: Benchmarking Machine Learning Agents for Scientific Research
  6. 2026MARS: Modular Agent with Reflective Search for Automated AI Research
  7. 2026DeltaML-Bench: Evaluating Machine Learning Agents on Real-World Research Repositories
  8. 2025MLRC-Bench: Can Language Agents Solve Machine Learning Research Challenges?