← Back to all news
arXiv Machine Learning By Alejandro Calatrava, Xu Zhang, Ren Wang

SAM-on-the-Curve: Sharpness-Aware Mode Connectivity for Robust Weight-Space Interpolation

Read the original on arXiv Machine Learning →

The paper introduces Sharp Mode Connectivity (SMC), a new formulation of mode connectivity that optimizes for flatness in the entire local neighborhood of a weight-space path rather than just along the path itself. By applying a sharpness-aware approximation to a minimax objective, SMC produces low-loss curves that remain robust to distribution shifts, achieving significant accuracy gains on CIFAR‑10‑C and outperforming standard mode connectivity on several architectures.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv Machine Learning.

Related stories

arXiv Machine Learning
Aug 27

Robust CurveMoE: Multi-Norm Adversarial Defense for Mixture-of-Experts Models via Mode Connectivity

Robust CurveMoE is a mixture‑of‑experts framework that protects neural networks against perturbations defined by multiple norm constraints. It connects norm‑specialized models through a low‑loss path, selectively expertises only influential layers, and shares the rest of the parameters across routing paths. The method introduces contribution‑guided partial updating to reduce curve‑construction cost and provides a theoretical bound on the objective gap between partial and full optimization, achieving consistent improvements in clean, norm‑specific, and Union accuracy on CIFAR‑100 and ImageNet‑100.

By Xu Zhang, Ren Wang
llmscomputer-visionsafety
More like this →
arXiv Computer Vision
Sep 15

Sparsity-Adaptive Sharpness-Aware Minimization

The paper introduces Sparsity-Adaptive Sharpness-Aware Minimization (SA‑SAM), a method that adjusts the perturbation radius in sharpness-aware training to remain consistent as model sparsity increases. It also evaluates a Magnitude‑Weighted Hessian (MWH) importance metric derived from second‑order analysis. Experiments on CIFAR‑10‑C, CIFAR‑100‑C, and ImageNet‑100‑C show that SA‑SAM improves corruption robustness at 80–90% sparsity while maintaining clean accuracy, and the study reports inference throughput at deployment‑relevant sparsity levels.

By Shiryu Ueno, Yoshikazu Hayashi, Kunihito Kato
efficiency
More like this →
arXiv Machine Learning
Jul 10

LiST: Lipschitz Scaling Training for Robust and Calibrated Neural Networks

arXiv:2607. 07745v1 Announce Type: new Abstract: While accuracy, robustness, and calibration are all essential for reliable neural networks, they are often studied separately; developing models that satisfy all three simultaneously remains a central challenge.

By Arthur Chiron (IRIT, EPE UT), Franck Mamalet (IRIT, DTIPG - SNCF, UT3), Thomas Massena (IRIT, DTIPG - SNCF, UT3), Thomas Deltort (IRIT), Mathieu Serrurier (IRIT, UT2J)
benchmarks
More like this →