This paper provides the first non-vacuous generalization analysis for the Stochastic Variance Reduced Gradient (SVRG) method by establishing sharp, data-dependent algorithmic stability bounds, thereby…

The paper introduces a Jacobian-based spectral audit to evaluate neural operators, demonstrating that standard prediction error metrics fail to capture crucial local dynamical structures and operator…

The paper introduces FoLoRA, a novel optimization framework that uses a generalized Rayleigh quotient to achieve a superior balance between adapting foundation models to specific tasks and preserving…

The paper introduces Sparse Memory-Efficient Training (SMET), a method that stabilizes and optimizes Dynamic Sparse Training (DST) for large language models, enabling stable and memory-efficient spars…

The paper introduces a history-aware adaptive Reduced-Order Model (ROM) framework using incremental Singular Value Decomposition (iSVD) that maintains accuracy for online dynamics far beyond the initi…

The paper introduces Singularity-aware Adam (S-Adam), a novel optimizer that stabilizes deep learning training in non-smooth loss landscapes by dynamically damping updates based on local geometric ins…

The paper analyzes a new class of asynchronous adaptive first-order optimization methods and proves their stochastic convergence rate is O(1/sqrt{t}) for non-convex functions.

TailLoR is a new parameter-efficient finetuning method that uses the singular bases of pre-trained weights to learn low-rank updates, specifically penalizing updates along dominant directions to impro…

The paper develops a quantitative framework to analyze and improve flow distillation in diffusion models, providing stability guarantees and suggesting non-uniform time scheduling to reduce approximat…

The paper proposes using a Physics-Informed Neural Network (PINN) residual as an efficient, physics-guided indicator to guide adaptive mesh refinement (AMR) for classical finite-difference PDE solvers…

The paper introduces a novel diffusion posterior sampling method that stabilizes and accelerates data-consistent sampling by replacing hand-tuned guidance weights with a per-noise-level, curvature-gui…

The paper systematically characterizes column-level activation sparsity across various diffusion model architectures, demonstrating that element-level sparsity metrics significantly overestimate the a…

The paper introduces SB-ECC, a novel score-based decoder that models error correction as continuous-time denoising, achieving state-of-the-art performance across various code families and noise levels…

STAB is a novel specification-driven pipeline that generates test cases exposing algorithmic bottlenecks by combining constraint-bound maximization and adversarial structure injection, significantly i…

DASH introduces a dual-branch distillation framework to effectively compress class-conditional diffusion models by independently supervising both score branches, significantly preserving guidance fide…

The paper proposes DiReCT, a novel framework that treats data selection during LLM annealing as a constrained optimization problem based on the spectral geometry of the loss landscape, achieving state…

This paper provides a comprehensive generalization analysis of Stochastic Gradient Descent with Momentum (SGDM) by establishing tight, on-average model stability bounds that show SGDM can generalize w…