AdaGram applies a dynamical low-rank integrator to the preconditioner of adaptive gradient methods, achieving full-matrix-like updates with rank 1-5 approximations at low cost on small GLM benchmarks.
SOAP: One-shot pruning of generative pre-trained transformer checkpoints with a second-order method, 2024
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
unclear 1representative citing papers
citing papers explorer
-
Dynamic Low-rank Approximation of Full-Matrix Preconditioner for Training Generalized Linear Models
AdaGram applies a dynamical low-rank integrator to the preconditioner of adaptive gradient methods, achieving full-matrix-like updates with rank 1-5 approximations at low cost on small GLM benchmarks.