Agent swarms speed up discovery and improve math problem solutions
SwarmEvolve: Latent Agent Swarms for Evolutionary Discovery and Optimization
Artificial IntelligenceNeural and Evolutionary Computing
Summary
Finding better program solutions for math and science problems can be very slow. The authors created a system where hundreds of simple AI agents work together in secret code form rather than text. These agents share ideas quickly, keep only the most unique ones, and combine their knowledge like a team brainstorming session. This new method solved challenging problems faster and better than previous AI approaches. It shows that teamwork inside hidden computer states can make scientific problem-solving more efficient.
What this means in practice
- •For software engineers: Boost the speed and efficiency of automated program optimization for math and scientific computing tasks by using latent-space agent swarms.
- •For data science teams: Improve rapid discovery of novel computational methods in scientific data analysis through collaborative AI agents sharing latent information.
Authors
Qile Jiang, George Em Karniadakis
Abstract
LLM-guided evolutionary search uses AI agents to improve programs for mathematical and computational problems by iteratively refining solutions and combining high-scoring methods. However, the process is often slow because the LLM generates a long textual chain of thought for each round and must repeat this generation many times to maintain population diversity. We present SwarmEvolve, a framework in which hundreds of low-cost agents per round work in swarms, reason in latent space without decoding to text, and share information by passing KV caches. Each swarm studies the problem from a different angle. A novelty filter keeps the most distinct agent ideas by directly measuring the distances between the agents' hidden state vectors. Crossover of ideas happens naturally in latent space, as a summarizer agent receives the combined KV caches of all swarm members. On 21 problems in mathematical discovery and scientific computing, SwarmEvolve gets the highest average score and needs a median of only 3 to 8 minutes to reach the best scores that other state-of-the-art frameworks reach in 60 minutes. On several problems, SwarmEvolve with Qwen3 8B improves on the solutions that AlphaEvolve first found with frontier LLMs. Our work shows that agent swarms that communicate in latent space offer a fast, low-cost approach to program search for scientific problems.