Reproduce “AgentWebRec: Compact Evidence Fusion over the Agent Web for Personalized Recommendation” on OpenResearch

Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.

Universal app for Apple silicon and Intel. Requires macOS 11 or later.

About this paper

Haoran Qiang, Guannan Liu, Liang Zhang, Junjie Wu

AgentWebRec assembles evidence for recommendation decisions rather than pooling users’ histories into a shared profile. It enriches item descriptions, retrieves relevant target-user memories, and asks neighboring agents for preference summaries when local confidence is low, while memories remain local. On Books, its H@1 is 0.5250 versus AgentCF’s 0.4850; this supports ranking under the benchmark test protocol, not deployment performance.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.