attention masking 1privileged information 1self-distillation 1sequential recommendation 1teacher-student learning 1
From the 1 of 2 linked papers with an AI index.
2 papers
cs.LG2026
Question Begets Question: Self-Evolving Curriculum for Reinforcement Fine-Tuning on Competition Mathematics
Longtian Bao, Jianyou Wang, Yang Zhang +2
Teaching a language model a skill it has not mastered is obstructed by three recurring difficulties: training data is scarce, ground-truth reasoning traces are usually unavailable,…
cs.IR2026
Learning from the Future: Privileged Self-Distillation for Sequential Recommendation
Jiakai Tang, Yang Zhang, See-Kiong Ng +4
The paper introduces Privileged Self-Distillation (PSD), a method that uses future user interactions as training‑only privileged information to improve sequential recommendation mo…