新闻
The Silent Cost of Conversation History in LLM Reasoning
arXiv cs.AI · 发布于 · 阅读约3分钟
30秒读懂
- 发生了什么
- Research shows conversation history degrades LLM reasoning accuracy over multi-turn exchanges, with accuracy dropping from 0.6 to 0.1 within twenty turns.
- 为何重要
- Engineers building multi-turn LLM systems should care when designing state management and conversation handling for production applications.
- 注意
- Early warning metrics tested did not reliably predict accuracy drops across models, limiting ability to detect degradation before it occurs.
- llm
- language model
- reasoning
- gpt
- llama
这条新闻背后的模式
每个模式都讲清楚技术如何运作、何时值得投入,以及在哪里会失效。
The Agent Architect
每周一个模式、一个权衡、一个生产事故案例。为构建智能体系统的人准备的每周简报。
每周一封邮件,一键退订。您的地址仅用于发送简报。