Kuaishou's UAGR injects utility signals into semantic codes and decoding for generative recommendation
_reachsumit · x · 2026-11-04
Kuaishou researchers propose UAGR, aligning both semantic ID (SID) construction and beam-search decoding with downstream multi-objective utility in generative recommendation.
The problem: conventional SIDs capture semantic similarity but underrepresent item utility from post-exposure behaviors like shares, long views, and video completion, and generation likelihood can prune high-utility candidates before reranking can compensate.
The fix: Utility-Enriched Semantic Codes embed item-level utility without increasing SID length or vocabulary size, supporting incremental updates with no re-tokenization cost; Utility-Guided Decoding estimates candidate utility during beam search to stop valuable candidates being dropped early.
More from Research
- François Chollet: modeling brain wiring as an engineering problem constrained by genome size and time — CatAstro_Piyush · 2026-11-04
- OffsecCyber dataset combines HackerOne reports and CVEs to train pentesting LLMs — AccBalanced · 2026-11-04
- Monofact rate in training data governs LLM scientific discovery, preprint finds — CatAstro_Piyush · 2026-11-04
- qwenTPU compiles a TPU into Qwen3.5 weights, achieving exact integer matmul — CatAstro_Piyush · 2026-11-04
- Jev model improves ICL demonstration retrieval via distillation into a lightweight MLP — kalyan_kpl · 2026-11-04
- Indie researcher open-sources 10 years of math work: 146k files, AI-accelerated over 3 years — HamOnWhy · 2026-11-04