51 citations · 96 across the 7 of their papers we have counts for
3 papers · 1 filter
Stepping on the Edge: Curvature Aware Learning Rate Tuners
Vincent Roulet, Atish Agarwala, Jean-Bastien Grill +3
Curvature information -- particularly, the largest eigenvalue of the loss Hessian, known as the sharpness -- often forms the basis for learning rate tuners. However, recent work ha…
Direct Language Model Alignment from Online AI Feedback
Shangmin Guo, Biao Zhang, Tianlin Liu +9
Direct alignment from preferences (DAP) methods, such as DPO, have recently emerged as efficient alternatives to reinforcement learning from human feedback (RLHF), that do not requ…
Routers in Vision Mixture of Experts: An Empirical Study
Tianlin Liu, Mathieu Blondel, Carlos Riquelme +1
Mixture-of-Experts (MoE) models are a promising way to scale up model capacity without significantly increasing computational cost. A key component of MoEs is the router, which dec…