团队协作的权重：通信结构能否帮助 LLM 科学家寻找科学定律？
原标题：The Weight of Teamwork: Does Communication Structure Help LLM Scientists Search for Scientific Laws?
AI 摘要
该论文以符号回归作为可观测的科学搜索任务，在固定调用预算下比较了不同通信结构（Independent、Ring、Star、Complete）的 LLM 多智能体团队，并分别评估其候选表达式存档与最终选出的表达式。结果显示，将预算分散到更多独立轨迹带来的提升最大（平均留出拟合分数提高 0.70），而 Ring 策略在统一存档选择器下比 Independent 低 0.33；Star 与 Complete 的差异尚不明确。作者据此主张应将候选生成与最终选择作为两个独立指标报告，并指出在最终选择规则未指定时，仅凭通信结构无法确定架构排名。
正文节选
The Weight of Teamwork: Does Communication Structure Help LLM Scientists Search for Scientific Laws? Abstract Team performance is often reported as one delivered score, even though that score is produced in two stages: a system generates candidate answers and a selection rule returns one of them. We use symbolic regression as an observable scientific-search setting and compare fixed-budget LLM teams that differ in peer visibility and update timing, evaluating both their frozen candidate archives