![[2602.16928] Discovering Multiagent Learning Algorithms with Large Language Models](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2602.16928] Discovering Multiagent Learning Algorithms with Large Language Models
Summary
This paper explores the use of large language models to automatically discover new multiagent learning algorithms, enhancing the efficiency of Multi-Agent Reinforcement Learning (MARL) in imperfect-information games.
Why It Matters
The research addresses the limitations of manual algorithm design in MARL by introducing AlphaEvolve, which leverages large language models to innovate algorithmic strategies. This advancement could significantly improve the performance and adaptability of AI systems in complex environments, making it relevant for both academic research and practical applications in AI.
Key Takeaways
- AlphaEvolve uses large language models to automate the discovery of multiagent learning algorithms.
- The paper presents two novel algorithms: VAD-CFR and SHOR-PSRO, which outperform existing methods.
- The research highlights the potential of AI to navigate complex algorithmic design spaces without human intervention.
- Innovative mechanisms like volatility-sensitive discounting and hybrid meta-solvers are introduced.
- This approach could lead to more efficient and effective AI systems in game-theoretic scenarios.
Computer Science > Computer Science and Game Theory arXiv:2602.16928 (cs) [Submitted on 18 Feb 2026] Title:Discovering Multiagent Learning Algorithms with Large Language Models Authors:Zun Li, John Schultz, Daniel Hennes, Marc Lanctot View a PDF of the paper titled Discovering Multiagent Learning Algorithms with Large Language Models, by Zun Li and 3 other authors View PDF HTML (experimental) Abstract:Much of the advancement of Multi-Agent Reinforcement Learning (MARL) in imperfect-information games has historically depended on manual iterative refinement of baselines. While foundational families like Counterfactual Regret Minimization (CFR) and Policy Space Response Oracles (PSRO) rest on solid theoretical ground, the design of their most effective variants often relies on human intuition to navigate a vast algorithmic design space. In this work, we propose the use of AlphaEvolve, an evolutionary coding agent powered by large language models, to automatically discover new multiagent learning algorithms. We demonstrate the generality of this framework by evolving novel variants for two distinct paradigms of game-theoretic learning. First, in the domain of iterative regret minimization, we evolve the logic governing regret accumulation and policy derivation, discovering a new algorithm, Volatility-Adaptive Discounted (VAD-)CFR. VAD-CFR employs novel, non-intuitive mechanisms-including volatility-sensitive discounting, consistency-enforced optimism, and a hard warm-start pol...