WorksComputer Science
LoopGR: Parameter-Efficient Scaling for Generative Recommendation via Recurrent Depth
Versions
Every version keeps its own PDF, citation, and review record — endorsements are bound to the version they examined and never carry forward silently.