SwarmEvolve：用于进化发现与优化的潜在智能体集群
SwarmEvolve: Latent Agent Swarms for Evolutionary Discovery and Optimization
Qile Jiang · George Em Karniadakis
中文摘要
LLM 引导的进化搜索利用 AI 智能体改进数学与计算问题的程序，通过迭代精炼解并组合高分方法。但该过程通常较慢，因为 LLM 每轮生成很长的文本思维链，且需重复多次以维持种群多样性。SwarmEvolve 框架让每轮数百个低成本智能体以集群方式工作，在潜在空间（latent space）中推理而不解码为文本，并通过传递 KV 缓存共享信息。每个集群从不同角度研究问题。新颖性过滤器直接测量智能体隐藏状态向量之间的距离，保留最独特的智能体想法。想法在潜在空间中自然交叉，因为汇总智能体接收所有集群成员组合后的 KV 缓存。在数学发现与科学计算的 21 个问题上，SwarmEvolve 取得最高平均分，达到其他最先进框架 60 分钟才取得的最佳分数所需时间中位数仅为 3 至 8 分钟。在若干问题上，使用 Qwen3 8B 的 SwarmEvolve 改进了 AlphaEvolve 最初用前沿 LLM 找到的解。工作表明，在潜在空间中通信的智能体集群为科学问题的程序搜索提供了快速、低成本的方法。
关键要点
- 01LLM 进化搜索因每轮生成长文本思维链且需重复多次而速度慢
- 02数百低成本智能体在潜在空间推理，通过传递 KV 缓存共享信息
- 03新颖性过滤器用隐藏状态向量距离保留最独特想法，交叉在潜在空间自然发生
- 0421 个问题上平均分最高，达最佳分数中位数仅 3 至 8 分钟，对比其他框架 60 分钟
- 05Qwen3 8B 在若干问题上改进 AlphaEvolve 用前沿 LLM 找到的解
解读
尚无解读。
原始英文摘要
arXiv:2611.08483v1 Announce Type: new 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.