Internalizing Multi-Agent Reasoning for Accurate and Efficient LLM-based Recommendation

ACL ARR 2026 January Submission9940 Authors

Keywords: Large Language Models, Recommender Systems, Multi-Agent Systems, Knowledge Distillation

Abstract: Large Language Models (LLMs) are reshaping recommender systems by leveraging extensive world knowledge and semantic reasoning to interpret user intent. However, effectively integrating these capabilities with collaborative signals while avoiding prohibitive inference latency remains a critical bottleneck. To address this, we propose a trajectory-driven internalization framework to develop a Single-agent Trajectory-Aligned Recommender (STAR). Specifically, to internalize complex reasoning capabilities into a single efficient model, we first design a multi-agent teacher system capable of multi-turn tool usage and reflection. This teacher utilizes a Collaborative Signal Translation mechanism to explicitly convert latent behavioral patterns into descriptive natural language evidence to enhance reasoning accuracy. Subsequently, a trajectory-driven distillation pipeline transfers this agentic logic, including planning, tool usage, and self-reflection, into the compact STAR model. Extensive experiments demonstrate that STAR surpasses its teacher by 8.7\% to 39.5\% while eliminating iterative latency, paving the way for real-time, reasoning-enhanced recommendation.

Paper Type: Long

Research Area: NLP Applications

Research Area Keywords: NLP Applications

Languages Studied: English

Submission Number: 9940