CORAL: Towards Autonomous Multi-Agent Evolution for Open-Ended Discovery

2026-04-02

AI Summary

Current large language model–driven evolutionary approaches rely on fixed heuristics and hard-coded rules, limiting their capacity for autonomous exploration in open-ended discovery tasks. This work proposes the first autonomous multi-agent evolutionary framework tailored for open-ended problem solving, enabling long-term collaboration, reflection, and robust operation without hard-coded rules through mechanisms such as persistent shared memory, asynchronous execution, heartbeat-based intervention, isolated workspaces, and decoupled evaluators. The method establishes new state-of-the-art results across ten diverse tasks in mathematics, algorithm design, and system optimization, achieving 3–10× higher efficiency with fewer evaluations. Notably, on Anthropic’s core benchmark, it reduces the cycle count required to discover optimal solutions from 1,363 to 1,103.

Abstract

Large language model (LLM)-based evolution is a promising approach for open-ended discovery, where progress requires sustained search and knowledge accumulation. Existing methods still rely heavily on fixed heuristics and hard-coded exploration rules, which limit the autonomy of LLM agents. We present CORAL, the first framework for autonomous multi-agent evolution on open-ended problems. CORAL replaces rigid control with long-running agents that explore, reflect, and collaborate through shared persistent memory, asynchronous multi-agent execution, and heartbeat-based interventions. It also provides practical safeguards, including isolated workspaces, evaluator separation, resource management, and agent session and health management. Evaluated on diverse mathematical, algorithmic, and systems optimization tasks, CORAL sets new state-of-the-art results on 10 tasks, achieving 3-10 times higher improvement rates with far fewer evaluations than fixed evolutionary search baselines across tasks. On Anthropic's kernel engineering task, four co-evolving agents improve the best known score from 1363 to 1103 cycles. Mechanistic analyses further show how these gains arise from knowledge reuse and multi-agent exploration and communication. Together, these results suggest that greater agent autonomy and multi-agent evolution can substantially improve open-ended discovery. Code is available at https://github.com/Human-Agent-Society/CORAL.

Problem

• open-ended discovery
• autonomous agents
• multi-agent evolution
• LLM-based evolution
• exploration heuristics

Innovation

• autonomous multi-agent evolution
• open-ended discovery
• persistent memory
• asynchronous execution
• LLM-based evolution