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…

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 a unified, constrained optimization framework using KL divergence and likelihood constraints to achieve effective and principled unlearning in diffusion models.

The paper introduces trust functions to filter weak supervision labels, enabling near-lossless weak-to-strong generalization by selectively training a strong student using only the most reliable weak…

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…

LoopFM proposes a novel framework to significantly improve knowledge distillation for recommendation systems by structuring the rich intermediate embeddings of large foundation models as input feature…

This paper investigates the phenomenon of 'copying' in Distribution Matching Distillation (DMD), finding that high-dimensional distillation causes student models to spontaneously reproduce the teacher…

The paper proposes FRA-Attack, a frequency-domain regularization method, to significantly improve the transferability of adversarial attacks against closed-source Multimodal Large Language Models (MLL…

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 scaling exponent in neural scaling laws is not fixed but systematically depends on the optimizer used, with preconditioned optimizers generally yielding steeper scaling.

GradSentry introduces a novel backdoor sample filtering method that uses the spectral entropy of individual sample gradients to detect poisoned data during LLM fine-tuning, proving effective even at h…

The paper introduces TRACER, a novel regularization framework that uses Weighted Moving Average (WMA) distillation to robustly finetune multimodal models, mitigating catastrophic forgetting and improv…

The paper introduces a novel, transferable learned attack (LT-MIA) that detects a universal 'signature of memorization' in language models, achieving high accuracy across diverse model architectures (…

The paper introduces a comprehensive benchmark to test if physics foundation models learn generalizable dynamics, finding that their performance is highly conditional and not universally general.

The paper introduces Score Broadcast and Decorrelation (SBD), a general theoretical framework that unifies broadcast-based credit assignment across various differentiable loss functions by leveraging…

This paper analyzes the poor performance of Meta-learning for Training-data Selection (MTS) and proposes that increasing the batch size and incorporating informative features can significantly improve…

The paper introduces and analyzes several novel data appraisal metrics, including the Vendi Score and matrix spectral functions, demonstrating that efficient optimization techniques make these metrics…