State of NLP Optimization

3 papers · avg viability 3.7

View topic page

Top papers

SympFormer: Accelerated attention blocks via Inertial Dynamics on Density Manifolds(5.0)
Gradients Must Earn Their Influence: Unifying SFT with Generalized Entropic Objectives(4.0)
Efficient Reasoning at Fixed Test-Time Cost via Length-Aware Attention Priors and Gain-Aware Training(2.0)