L-SR1 is a learned second-order optimizer whose rank-one preconditioner is trained with a secant penalty, and it outperforms LGD on 3DPW human mesh recovery while using a smaller model.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
L-SR1: Learned Symmetric-Rank-One Preconditioning
L-SR1 is a learned second-order optimizer whose rank-one preconditioner is trained with a secant penalty, and it outperforms LGD on 3DPW human mesh recovery while using a smaller model.