记忆架构比通道容量更关键，决定大模型如何从零创造共享语言
From Signals to Structure: How Memory Architecture Drives Language Emergence in LLM Agents
- 用带私有笔记的内存让智能体持续积累沟通规则
- 笔记型架构在25通道下稳定协调率达86.7%，优于无记忆模型
- 适合研究智能体协作、语言演化与记忆机制的学者
在刘易斯信号博弈中，发送方和接收方需仅凭交互历史构建共享代码。我们研究了五种不同记忆架构在多种通道配置下的表现，发现记忆架构的影响大于通道容量。带有持久私有笔记的智能体能充分利用冗余通道容量，避免无状态智能体在高容量时的性能崩溃，在通道容量为25时达到最可靠协调（0.867 ± 0.023）。无状态智能体在中等容量时表现最佳，随后随词汇量增长而退化，因滚动上下文窗口无法追踪全部信息。笔记机制将已学惯例外部化，使智能体无需每轮重推代码。尽管信息瓶颈理论预测最优容量等于对象数，但容量为8时反而成为脆弱点，冗余容量通常更优。结果表明，仅靠通道容量无法预测协调效果；记忆架构决定了智能体能否将交互历史转化为稳定惯例，二者共同决定信号如何演变为语言。
原文摘要 · Abstract (English)
How do two agents invent a shared language from scratch? In a Lewis signaling game, a sender and receiver must coordinate on a code using only their interaction history. We study five memory architectures across varying channel configurations with LLM agents and find that memory architecture matters more than channel capacity. Agents with a persistent private notebook benefit from surplus channel capacity and avoid the high-capacity collapse seen in stateless agents, achieving the most reliable coordination ($0.867 \pm 0.023$ at capacity = 25). Stateless agents peak at moderate capacity and then degrade as the vocabulary grows beyond what a rolling context window can track The notebook externalizes learned conventions, freeing agents from having to re-derive codes each round. An information bottleneck-inspired argument predicts an optimal capacity equal to the number of objects. Instead, the bottleneck (capacity = 8) proves to be a fragility point, and surplus capacity is generally better. We show that channel capacity alone cannot predict coordination; memory architecture determines whether agents turn interaction history into stable conventions, and both dimensions are needed to understand how signals become language.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。