MemRec: Collaborative Memory-Augmented Agentic Recommender System

TL;DR

MemRec tackles the bottleneck of leveraging collaborative signals in agentic recommender systems by architecturally decoupling reasoning from memory management. It builds a dynamic collaborative memory graph managed by a dedicated LM_MEM, enabling efficient retrieval, synthesis, and asynchronous propagation of high-signal signals to a downstream LLM_REC. Through a three-stage pipeline—Collaborative Memory Retrieval, Grounded Reasoning, and Asynchronous Propagation—MemRec achieves state-of-the-art results on four benchmarks and establishes a Pareto-optimal trade-off between reasoning quality, cost, and deployment constraints. The work demonstrates strong empirical gains, rigorous ablations, and flexible deployment options, including local and cloud-native configurations, while outlining avenues for scalability and privacy-preserving updates.

Abstract

The evolution of recommender systems has shifted preference storage from rating matrices and dense embeddings to semantic memory in the agentic era. Yet existing agents rely on isolated memory, overlooking crucial collaborative signals. Bridging this gap is hindered by the dual challenges of distilling vast graph contexts without overwhelming reasoning agents with cognitive load, and evolving the collaborative memory efficiently without incurring prohibitive computational costs. To address this, we propose MemRec, a framework that architecturally decouples reasoning from memory management to enable efficient collaborative augmentation. MemRec introduces a dedicated, cost-effective LM_Mem to manage a dynamic collaborative memory graph, serving synthesized, high-signal context to a downstream LLM_Rec. The framework operates via a practical pipeline featuring efficient retrieval and cost-effective asynchronous graph propagation that evolves memory in the background. Extensive experiments on four benchmarks demonstrate that MemRec achieves state-of-the-art performance. Furthermore, architectural analysis confirms its flexibility, establishing a new Pareto frontier that balances reasoning quality, cost, and privacy through support for diverse deployments, including local open-source models. Code:https://github.com/rutgerswiselab/memrec and Homepage: https://memrec.weixinchen.com

Figures (16)

• Figure 1: (a) Existing Agents interact with user and item memories through separate, isolated read/write channels. (b) MemRec performs collaborative operations on memory graph, enabling global connectivity.
• Figure 2: The overall framework of MemRec, decoupling reasoning ($\text{LLM}_{\text{Rec}}$) from memory management ($\text{LM}_{\text{Mem}}$). The three-stage pipeline consists: Collaborative Memory Retrieval, synthesizing high-order connectivity context from memory graph; Grounded Reasoning, scoring items based on instruction and context; and Asynchronous Collaborative Propagation, evolving the semantic memory graph in the background.
• Figure 3: Impact of architectural decoupling on H@1. MemRec (blue) overcomes the information bottleneck that causes Naive Agents (orange) to plateau, achieving substantial gains over both Naive and Vanilla LLM (gray).
• Figure 4: Efficiency-Cost-Performance Landscape across LLM-based approaches. This bubble chart visualizes the trade-offs between reasoning performance (H@1), estimated computational cost, and sequential latency (bubble size). The dashed line marks the new Pareto frontier established by MemRec variants (blue), demonstrating superior trade-offs compared to simple LLM baselines (gray) and competing agents (orange).
• Figure 5: Rationale Quality Evaluation (GPT-4o Judge, 1-5 scale). Error bars show 95% CIs; *** denotes $p<0.001$, while ns means not significant on paired t-test.