Seeing and Reflecting: Multimodal Memory-Enhanced Agent Collaboration for Recommendation
For recommender systems, this framework addresses the limitations of text-centric inputs and coarse memory updates in LLM-based agents, improving multimodal reasoning and robustness.
MMEACR introduces a dual-track memory architecture for LLM-based agentic recommenders, separating reasoning from multimodal matching to handle visual evidence and preference drift. It achieves strong performance across three domains, with notable gains in visually grounded scenarios.
Large language model (LLM)-based agentic recommender systems show promise in modeling user preferences through natural-language reasoning, yet they remain limited by text-centric inputs and coarse-grained memory updates, making agents prone to missing visual evidence, semantic noise, and preference drift. To address these limitations, we propose MMEACR, a Multimodal Memory-Enhanced Agent Collaboration framework for recommendation. MMEACR introduces a dual-track memory architecture that separates interpretable agent reasoning from fine-grained multimodal matching. In the reasoning track, collaborative User and Item Memory Agents maintain persistent multimodal memories and update them through an attribute-guided reinforcement-and-reflection mechanism. In the matching track, a decoupled multi-modal embedding memory is built from raw interaction narratives and item images to preserve detailed cross-modal signals beyond structured memory updates. The two tracks are integrated through weighted Reciprocal Rank Fusion to produce robust and interpretable rankings. Experiments on three real-world domains show that MMEACR achieves strong overall performance against competitive LLM-based and agent-based baselines, with notable gains in visually grounded recommendation scenarios.