Reproduce “WhisperRec: Latent Reasoning for Efficient Foundation Recommendation 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

Hao Jiang, Peiru Du, Pengfei Yao, Mengting Li, Siyuan Lou, Kuo Cai, Sheng Yu, Qiang Luo and 5 more

WhisperRec predicts recommendation item codes using learned latent tokens trained with teacher-generated reasoning, rather than generating a rationale at inference. On Kuaishou’s industrial dataset, SID Hit@64 was 17.44% above OneReason’s Think variant and 9.33% above its No-Think variant. In a test of 872 instances, QPS was 17.49 times that of explicit CoT; this comparison does not isolate latent tokens’ contribution.

Summary by alphaXiv.

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