mlsanityData to 7 Oct 2026

AgentGA: Evolving Code Solutions in Agent-Seed Space

arXiv 2604.14655 ↗First posted 16 Apr 2026cs.AI cs.LGcited by 0

Not on the map. Why?

Abstract (arXiv)

We present AgentGA, a framework that evolves autonomous code-generation runs by optimizing the agent seed: the task prompt plus optional parent archives that initialize a fresh workspace. The outer loop searches over these reusable starting conditions rather than editing code directly. Each generation launches a fresh autonomous run in an isolated workspace, while selected parent archives provide inherited artifacts that descendants can inspect and reuse. AgentGA couples a population-level genetic algorithm with long-horizon agents; selection uses deterministic 1:1 elite tournaments and operator allocation is adapted online with a modified Hedge controller. We instantiate the approach for tabular AutoML on the 16-competition Weco-Kaggle Lite benchmark. Across the full benchmark, AgentGA averages 71.90% Exceeds % of Human versus 51.38% for the AIDE reference, winning 15/16 competitions. Within AgentGA runs, descendants conditioned on inherited parent archives win 51.9% of 1,680 parent-child tournaments versus 8.6% for de novo proposals. These results support agent-seed optimization as a practical design choice for autonomous code-search systems.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 23 of its 33 references that are in the map. First posted on arXiv 16 Apr 2026.

Builds on

23 of 33 references.

References · 23

  1. 2022Competition-Level Code Generation with AlphaCodecited by 1,427
  2. 2025AlphaEvolve: A coding agent for scientific and algorithmic discoverycited by 763
  3. 2020AutoGluon-Tabular: Robust and Accurate AutoML for Structured Datacited by 368
  4. 2025Darwin Godel Machine: Open-Ended Evolution of Self-Improving Agentscited by 294
  5. 2024ReEvo: Large Language Models as Hyper-Heuristics with Reflective Evolutioncited by 266
  6. 2024Data Interpreter: An LLM Agent For Data Sciencecited by 222
  7. 2025AIDE: AI-Driven Exploration in the Space of Codecited by 206
  8. 2023EvoPrompt: Connecting LLMs with Evolutionary Algorithms Yields Powerful Prompt Optimizerscited by 189
  9. 2020Auto-Sklearn 2.0: Hands-free AutoML via Meta-Learningcited by 174
  10. 2024AgentSquare: Automatic LLM Agent Search in Modular Design Spacecited by 141
  11. 2023Large Language Models for Automated Data Science: Introducing CAAFE for Context-Aware Automated Feature Engineeringcited by 125
  12. 2024DS-Agent: Automated Data Science by Empowering Large Language Models with Case-Based Reasoningcited by 119
11 more
  1. 2024AutoML-Agent: A Multi-Agent LLM Framework for Full-Pipeline AutoMLcited by 84
  2. 2016Evaluation of a Tree-based Pipeline Optimization Tool for Automating Data Sciencecited by 73
  3. 2024AutoKaggle: A Multi-Agent Framework for Autonomous Data Science Competitionscited by 70
  4. 2023MLCopilot: Unleashing the Power of Large Language Models in Solving Machine Learning Taskscited by 66
  5. 2025SE-Agent: Self-Evolution Trajectory Optimization in Multi-Step Reasoning with LLM-Based Agentscited by 66
  6. 2025EvoFlow: Evolving Diverse Agentic Workflows On The Flycited by 49
  7. 2026Group-Evolving Agents: Open-Ended Self-Improvement via Experience Sharingcited by 32
  8. 2024SELA: Tree-Search Enhanced LLM Agents for Automated Machine Learningcited by 23
  9. 2024Kolb-Based Experiential Learning for Generalist Agents with Human-Level Kaggle Data Science Performancecited by 22
  10. 2025MLZero: A Multi-Agent System for End-to-end Machine Learning Automationcited by 22
  11. 2025I-MCTS: Enhancing Agentic AutoML via Introspective Monte Carlo Tree Searchcited by 14

Built on by

Not cited yet.