Showing cs.LGShow all
2 papers · 1 filter
cs.LG2026
Hidden Failure Modes of Gradient Modification under Adam in Continual Learning, and Adaptive Decoupled Moment Routing as a Repair
Yuelin Hu, Zhenbo Yu, Zhengxue Cheng +2
Many continual-learning methods modify gradients upstream (e.g., projection, penalty rescaling, replay mixing) while treating Adam as a neutral backend. We show this composition ha…
cs.LG2026
Entropy-Gated Selective Policy Optimization:Token-Level Gradient Allocation for Hybrid Training of Large Language Models
Yuelin Hu, Zhengxue Cheng, Wei Liu +1
Hybrid training methods for large language models combine supervised fine tuning (SFT) on expert demonstrations with reinforcement learning (RL) on model rollouts, typically at the…