mlsanityData to 14 Oct 2026

A Self-Improving Coding Agent

arXiv 2504.15228 ↗First posted 21 Apr 2025cs.AIcited by 79

On the map: LLM agents and tool use → Software engineering coding agents

also in 1 more
Abstract (arXiv)

Recent advancements in Large Language Models (LLMs) have spurred interest in deploying LLM agents to undertake tasks in the world. LLMs are often deployed in agent systems: code that orchestrates LLM calls and provides them with tools. We demonstrate that an agent system, equipped with basic coding tools, can autonomously edit itself, and thereby improve its performance on benchmark tasks. We find performance gains from 17% to 53% on a random subset of SWE Bench Verified, with additional performance gains on LiveCodeBench, as well as synthetically generated agent benchmarks. Our work represents an advancement in the automated and open-ended design of agentic systems, and demonstrates a data-efficient, non gradient-based learning mechanism driven by LLM reflection and code updates.

Key facts. Cited by 79 papers in the map (data through 14 Oct 2026). Placed in Software engineering coding agents, one of 16 directions in LLM agents and tool use. Builds on 37 of its 56 references that are in the map. First posted on arXiv 21 Apr 2025.

Builds on

37 of 56 references.

References · 37

  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. 2020Measuring Massive Multitask Language Understandingcited by 7,963
  4. 2025DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learningcited by 6,330
  5. 2022Large Language Models are Zero-Shot Reasonerscited by 4,374
  6. 2023Reflexion: Language Agents with Verbal Reinforcement Learningcited by 4,276
  7. 2023Toolformer: Language Models Can Teach Themselves to Use Toolscited by 4,182
  8. 2023Self-Refine: Iterative Refinement with Self-Feedbackcited by 3,442
  9. 2023Tree of Thoughts: Deliberate Problem Solving with Large Language Modelscited by 3,211
  10. 2023GPQA: A Graduate-Level Google-Proof Q&A Benchmarkcited by 2,879
  11. 2022Do As I Can, Not As I Say: Grounding Language in Robotic Affordancescited by 2,348
  12. 2023Voyager: An Open-Ended Embodied Agent with Large Language Modelscited by 2,191
25 more
  1. 2024LiveCodeBench: Holistic and Contamination Free Evaluation of Large Language Models for Codecited by 2,004
  2. 2023Improving Factuality and Reasoning in Language Models through Multiagent Debatecited by 1,784
  3. 2022Program of Thoughts Prompting: Disentangling Computation from Reasoning for Numerical Reasoning Taskscited by 1,076
  4. 2022STaR: Bootstrapping Reasoning With Reasoningcited by 940
  5. 2019DROP: A Reading Comprehension Benchmark Requiring Discrete Reasoning Over Paragraphscited by 903
  6. 2023Graph of Thoughts: Solving Elaborate Problems with Large Language Modelscited by 886
  7. 2023Large Language Models as Optimizerscited by 864
  8. 2023Promptbreeder: Self-Referential Self-Improvement Via Prompt Evolutioncited by 551
  9. 2022Language Models are Multilingual Chain-of-Thought Reasonerscited by 530
  10. 2025The AI Scientist-v2: Workshop-Level Automated Scientific Discovery via Agentic Tree Searchcited by 422
  11. 2024Automated Design of Agentic Systemscited by 407
  12. 2024Recursive Introspection: Teaching Language Model Agents How to Self-Improvecited by 165
  13. 2024AgentSquare: Automatic LLM Agent Search in Modular Design Spacecited by 155
  14. 2023Self-Taught Optimizer (STOP): Recursively Self-Improving Code Generationcited by 148
  15. 2025MLGym: A New Framework and Benchmark for Advancing AI Research Agentscited by 124
  16. 2024Secret Collusion among AI Agents: Multi-Agent Deception via Steganographycited by 113
  17. 2023Motif: Intrinsic Motivation from Artificial Intelligence Feedbackcited by 102
  18. 2020Autotelic Agents with Intrinsically Motivated Goal-Conditioned Reinforcement Learning: a Short Surveycited by 85
  19. 2024Open-Endedness is Essential for Artificial Superhuman Intelligencecited by 80
  20. 2024OMNI-EPIC: Open-endedness via Models of human Notions of Interestingness with Environments Programmed in Codecited by 74
  21. 2023OMNI: Open-endedness via Models of human Notions of Interestingnesscited by 70
  22. 2022Parsel: Algorithmic Reasoning with Language Models by Composing Decompositionscited by 64
  23. 2024MaestroMotif: Skill Design from Artificial Intelligence Feedbackcited by 24
  24. 2024Practice Makes Perfect: Planning to Learn Skill Parameter Policiescited by 23
  25. 2024Multi-expert Prompting Improves Reliability, Safety, and Usefulness of Large Language Modelscited by 14

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 14; 2026: 65.

79 papers.

Sort:

By area:

  1. 2025Darwin Godel Machine: Open-Ended Evolution of Self-Improving Agentscited by 356
  2. 2025A Survey of Self-Evolving Agents: What, When, How, and Where to Evolve on the Path to Artificial Super Intelligencecited by 170
  3. 2025SWE-EVO: Benchmarking Coding Agents in Long-Horizon Software Evolution Scenarioscited by 59
  4. 2025Huxley-G"odel Machine: Human-Level Coding Agent Development by an Approximation of the Optimal Self-Improving Machinecited by 53
  5. 2026A Survey of Agentic Reasoning for Large Language Models: Towards Recursively Self-Improving and Collective Agentscited by 50
  6. 2026HarnessX: A Composable, Adaptive, and Evolvable Agent Harness Foundrycited by 47
  7. 2025Vibe Coding vs. Agentic Coding: Fundamentals and Practical Implications of Agentic AIcited by 42
  8. 2026Group-Evolving Agents: Open-Ended Self-Improvement via Experience Sharingcited by 36
  9. 2026Structured Scaling of AI Discovery Across Diverse Scientific Domainscited by 21
  10. 2025Magentic Marketplace: An Open-Source Environment for Studying Agentic Marketscited by 19
  11. 2026The Red Queen G"odel Machine: Co-Evolving Agents and Their Evaluatorscited by 19
  12. 2026Self-Improvements in Modern Agentic Systems: A Surveycited by 19
38 more
  1. 2026SimWorld Studio: Automatic Environment Generation with Evolving Coding Agent for Embodied Agent Learningcited by 17
  2. 2026MOSS: Self-Evolution through Source-Level Rewriting in Autonomous Agent Systemscited by 13
  3. 2026DarwinX: Evolving Agent Harnesses Through Natural Selectioncited by 13
  4. 2025Position: The Real Barrier to LLM Agent Usability is Agentic ROIcited by 12
  5. 2026From Question Answering to Task Completion: A Survey on Agent System and Harness Designcited by 12
  6. 2026The Last AI Built by Humans: Toward Genuine Recursive Self-Improvementcited by 11
  7. 2026Recursive self-improvement of AI research agentscited by 11
  8. 2026HarnessOpt-Bench: Evaluating LLMs at Harness Optimizationcited by 9
  9. 2026AutoDesign: Meta-Harness Optimization for Long-Horizon Agentic Designcited by 9
  10. 2026Recursive Experiential-Working Memory Evolution for Long-Horizon Agent Harnessescited by 9
  11. 2026Hierarchical Self-Improvement: A Framework for Task-Specific Evolvable Agent Harnessescited by 7
  12. 2025VisCoder2: Building Multi-Language Visualization Coding Agentscited by 6
  13. 2025DeliveryBench: Can Agents Earn Profit in Real World?cited by 6
  14. 2026DemoEvolve: Overcoming Sparse Feedback in Agentic Harness Evolution with Demonstrationscited by 6
  15. 2025Self-Correcting Code Generation Using Small Language Modelscited by 4
  16. 2025Exploring the Challenges and Opportunities of AI-assisted Codebase Generationcited by 4
  17. 2026Self Improvement via Fast Tree-searchcited by 4
  18. 2025When "Correct" Is Not Safe: Can We Trust Functionally Correct Patches Generated by Code Agents?cited by 3
  19. 2026Beyond Code Snippets: Benchmarking LLMs on Repository-Level Question Answeringcited by 3
  20. 2026Ouroboros: A Self-Developing Frontier Coding Agent with Reviewed Core Evolutioncited by 3
  21. 2026A Survey on Self-Improving Test-Time Intelligence: Feedback-Driven Adapting, Learning, and Scaling at Inferencecited by 3
  22. 2025Textual Gradients are a Flawed Metaphor for Automatic Prompt Optimizationcited by 2
  23. 2026Evolving Ensemble of Agentscited by 2
  24. 2026HiLSVA: Design and Evaluation of a Human-in-the-Loop Agentic System for Scientific Visualizationcited by 2
  25. 2026AI-Assisted Discovery of Convex Relaxations via Dual Agentscited by 2
  26. 2026Self-Evolving Coding Agentscited by 2
  27. 2026Mechanist: AI as a Scientific Instrument for Discovering the Mechanisms of Intelligencecited by 2
  28. 2026A Theory of Reliable Self-Evolution for Agent Harnessescited by 2
  29. 2025Self-evolving expertise in complex non-verifiable subject domains: dialogue as implicit meta-RLcited by 1
  30. 2026Early Discoveries of Algorithmist I: Promise of Provable Algorithm Synthesis at Scalecited by 1
  31. 20261GC-7RC: One Graphic Card -- Seven Research Challenges! How Good Are AI Agents at Doing Your Job?cited by 1
  32. 2026Lean4Agent: Formal Modeling and Verification for Agent Workflow and Trajectorycited by 1
  33. 2026Self-Modifying Lean Proof Agents with Verifier-Grounded Benchmark Coevolutioncited by 1
  34. 2026Eureka: Task-Conditioned Meta-Agent Orchestration for Scientific Discoverycited by 1
  35. 2026Agents in the Large: Perception-Centered Architecture for Persistent Agentscited by 1
  36. 2026Harness as a Language: A Minimalist Agent Framework With Maximal Expressivitycited by 1
  37. 2026Certified Long-Horizon Code Agent Evolution via Validation-Gated Skill Optimizationcited by 1
  38. 2026A Blueprint for Self-Evolving Coding Agents in Vehicle Aerodynamic Drag Predictioncited by 0

50 most cited of 79 All 79 (JSON)

Similar, not cited

Close in wording, no citation either way

  1. 2025SEW: Self-Evolving Agentic Workflows for Automated Code Generation
  2. 2025MapCoder-Lite: Distilling Multi-Agent Coding into a Single Small LLM
  3. 2024AgentQuest: A Modular Benchmark Framework to Measure Progress and Improve LLM Agents
  4. 2025Lita: Light Agent Uncovers the Agentic Coding Capabilities of LLMs
  5. 2026AgentGym2: Benchmarking Large Language Model Agents in De-Idealized Real-World Environments
  6. 2025AI Agentic Programming: A Survey of Techniques, Challenges, and Opportunities
  7. 2024Benchmark Self-Evolving: A Multi-Agent Framework for Dynamic LLM Evaluation
  8. 2023AgentBench: Evaluating LLMs as Agents