New routing method improves multi-turn AI conversations accuracy
SWRouter: Similarity-Contractive Window Routing for Multi-Turn Large Language Model Conversations
Computation and LanguageArtificial IntelligenceInformation Retrieval
Summary
Handling conversations with AI models that involve many turns is tricky because the system has to remember and use past dialogue correctly. The authors found that existing ways to choose which AI model answers a question work well for single questions but not for longer chats. They created SWRouter, a method that better organizes past dialogue to pick the best AI model for each new question. Their tests show this method improves accuracy by a significant amount compared to previous approaches.
What this means in practice
- •For conversational ai developers: Enhance multi-turn chatbot systems by routing user queries to the optimal language model using improved context segmentation.
- •For customer support teams: Improve automated multi-turn dialogue responses by integrating routing that better manages historical context for model selection.
Authors
Yu Wang, Yuchen Li, Rui Kong, Xinran Chen, Jiamin Chen, Hengyi Cai, Shuaiqiang Wang, Jiashu Zhao, Yulun Zhang, Zhonghao Lyu, Haoyi Xiong, Linghe Kong, Jimmy Xiangji Huang, Dawei Yin
Abstract
Large language models exhibit complementary strengths, motivating routing methods that dispatch each query to the most suitable model. Although existing routers are effective in single-turn settings, they do not directly transfer to multi-turn dialogue, where routing performance critically depends on how historical context is segmented, retained, and incorporated into the current prompt. This introduces two fundamental challenges: preventing information loss and information confusion during context construction, and evaluating routing quality without conflating model selection with prompt construction quality. In this paper, we propose SWRouter, a Similarity-Contractive Window Router for multi-turn large language model routing. SWRouter combines a similarity-based context segmentation mechanism for prompt construction with a dual-metric evaluation framework that decouples construction accuracy from router performance. Experiments on multi-turn dialogue benchmarks demonstrate that SWRouter consistently surpasses strong baselines, achieving a 16.26% improvement in evaluation accuracy over the best individual large language model and an additional 8.22% gain over the Conv-ID Context baseline. Our results highlight that multi-turn large language model routing requires a joint design of context construction and evaluation, rather than a direct extension of single-turn routing methods.