SOURCE-LINKED INTELLIGENCE

From Interests to Semantic IDs: Retrieval-Grounded Credit Assignment for Generative Recommendation

arXiv · AI, language, vision and robotics · article · Sep 24, 2026 · UTC

Semantic IDs (SIDs) encode each catalog item as a short token sequence, enabling generative recommenders to predict the next item autoregressively. Reasoning-enhanced variants, an increasingly common extension, first generate a textual trace and then decode a next-item SID by beam search. Such recommenders are commonly trained with group-relative policy optimization under an exact-match SID reward, which is sparse in large catalogs. Two failure modes follow. When all rollouts in a group miss the target, the group yields zero advantage and no learning signal. Rollouts sharing the same SID rewar

Read original source ↗ Open in workspace

recordType

paper

region

Global