Reproduce “RecGroupMemory: Feedback-Aware Cohort Memory for E-Commerce 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
Wendong Zhang, Weipeng Huang, Zhixin Ma, Bin Zhang, Dimin Wang
RecGroupMemory addresses online recommendation planning by letting an LLM revise cohort-level guidance with previous-day outcomes while preserving seven selected source values. The guidance feeds an orchestration LLM, not an item-scoring or ranking system. Among business-selected Taobao users with low purchasing power, Transaction UV rose 1.78% with feedback-aware memory versus its own no-memory baseline; this does not isolate feedback’s causal effect.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.