Reproduce “Decision-Native Recommendation with Large Language Models” 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
Daoxuan Liu, Jiayi Liao, Jiancan Wu, An Zhang, Xiang Wang, Xiangnan He
DICE selects the next item from a supplied candidate set by scoring each option against a user’s chronological history, rather than generating an item name. Its scoring head produces a probability distribution, with a tree layout reusing history computation. On Goodreads with 20 candidates, DICE-I reached HR@1 of 0.5332 versus 0.4718 for LlamaRec; tests cover candidate choice, not full-catalog retrieval.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.