TurboEvolve: Towards Fast and Robust LLM-Driven Program Evolution

arXiv:2604.18607 · cs.NE, cs.AI · Submitted 2026-04-12 · Read on arXiv

cs.NE, cs.AI

Submitted: 2026-04-12

Updated: 2026-10-12

Comments: 12 pages, 8 figures

Code: https://github.com/algorithmicsuperintelligence/openevolve

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Terminology

• Benchmark — 1× in this paper · explained in 2 episodes
• Robustness — 1× in this paper · explained in 5 episodes
• Sample Efficiency — 1× in this paper · explained in 4 episodes

Sources

• CodeEvolve: an open source evolutionary coding agent for algorithmic discovery and optimization
• Promptbreeder: Self-Referential Self-Improvement Via Prompt Evolution
• UniXcoder: Unified Cross-Modal Pre-training for Code Representation
• GigaEvo: An Open Source Optimization Framework Powered By LLMs And Evolution Algorithms
• ShinkaEvolve: Towards Open-Ended And Sample-Efficient Program Evolution
• Illuminating search spaces by mapping elites
• AlphaEvolve: A coding agent for scientific and algorithmic discovery
• LoongFlow: Directed Evolutionary Search via a Cognitive Plan-Execute-Summarize Paradigm
• ThetaEvolve: Test-time Learning on Open Problems
• Learning to Discover at Test Time
• Verbalized Sampling: How to Mitigate Mode Collapse and Unlock LLM Diversity

Related papers

• Evolutionary Ensemble of Agents
• Encoding and Decoding Temporal Signals with Spiking Bandpass Wavelets
• Large Language Models and Evolutionary Computation: A Critical Review of Bidirectional Interaction, Automated Algorithm Design, and Co-Adaptive Systems
• Evolutionary Architecture Search for Chlorophyll- a Prediction in Lakes using Sentinel-2
• Learning Alzheimer's Disease Signatures by bridging EEG with Spiking Neural Networks and Biophysical Simulations
• Neuromorphic Satellite Intelligence: A Formal Framework for Event-Driven Architectures, Spike-Based Algorithms, and Quantitative Energy-Accuracy Trade-offs