DARE reuses up to 87% of attention activations in diffusion LLMs through KV caching and output reuse, delivering 1.2x per-layer latency gains with average performance drops of 1.2-2.0%.
Training transformers with enforced lipschitz constants
7 Pith papers cite this work. Polarity classification is still indexing.
7
Pith papers citing it
citation-role summary
background 1
citation-polarity summary
years
2026 7roles
background 1polarities
background 1representative citing papers
Discrete tokenization in scientific foundation models imposes a geometric alignment tax that distorts continuous manifolds, with continuous heads reducing distortion by up to 8.5x and exposing three failure regimes in 14 biological models.
Deeper transformer layers produce intermediate representations that are harder to lossy-compress, and the paper links this to growing covariance and Rademacher complexity.
Splitting weight matrices into a fixed-norm direction and learnable per-row/column magnitudes improves LLM training over AdamW/Muon, removes weight decay and warmup, and transfers the optimal LR across width.
Entry-wise clipping achieves spectral control of gradients via localization under heavy-tailed contamination, with O(ε^{-4}) convergence and empirical savings on NanoGPT pretraining.
Muon optimizer in adversarial training imposes spectral-norm stability on matrix updates and matches or exceeds SGD/AdamW robustness on CNNs and ViTs under lp attacks.
Proximal stochastic spectral preconditioning converges for nonconvex constrained objectives under heavy-tailed noise, with a variance-reduced version achieving faster rates and a refined analysis of Muon iterations.
citing papers explorer
-
Constrained Stochastic Spectral Preconditioning Converges for Nonconvex Objectives
Proximal stochastic spectral preconditioning converges for nonconvex constrained objectives under heavy-tailed noise, with a variance-reduced version achieving faster rates and a refined analysis of Muon iterations.