AI不再写字，改用脑内电波开会
让 AI 改进算法，过去像让一群人在纸上写长信互相传：每轮都要把想法写成一大段文字，慢且贵。这篇让几百个小模型不再开口，直接在内部状态里「对脑电波」——把各自的记忆缓存互相传递，省掉写字这一步；谁的想法太像别人，就按内部向量的距离筛掉，保持多样性；不同思路的合体也不用写出来，一个汇总 agent 直接接收所有人的缓存，合并自然发生。效果：在 21 个数学与科学计算问题上，它平均分最高，拿到别人 60 分钟才达到的最好成绩，自己只要 3 到 8 分钟；在好几个问题上，它用 8B 的小模型就改进了 AlphaEvolve 用前沿大模型首次找到的解。它不是你明天能直接上手的东西，但它指向一件事：AI 群体协作的成本，可能不在算力，而在「非要说人话」。
📄 原文摘要(英文)
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.