SmartGR distills a large generative recommender into a smaller one with hierarchy-aware SID and beam-aware ranking losses, improving metrics by 8.6% on average while keeping the smaller model's speed.
International Conference on Learning Representations , year =
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.IR 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
SmartGR: Hierarchy and Beam-Aware Knowledge Distillation for Generative Recommendation
SmartGR distills a large generative recommender into a smaller one with hierarchy-aware SID and beam-aware ranking losses, improving metrics by 8.6% on average while keeping the smaller model's speed.