paper-with-me

홈 › Papers

Robustly Learning a Single Neuron via Sharpness

2023-06-13 · Puqian Wang, Nikos Zarifis, Ilias Diakonikolas, Jelena Diakonikolas

We study the problem of learning a single neuron with respect to the $L_2^2$-loss in the presence of adversarial label noise. We give an efficient algorithm that, for a broad family of activations including ReLUs, approximates the optimal $L_2^2$-error within a constant factor. Our algorithm applies under much milder distributional assumptions compared to prior work. The key ingredient enabling our results is a novel connection to local error bounds from optimization theory.

📄 PDF Abstract BibTeX arXiv:2306.07892

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

Robustly Learning Single-Index Models via Alignment Sharpness

2024-02-27 · Nikos Zarifis, Puqian Wang, Ilias Diakonikolas, Jelena Diakonikolas

We study the problem of learning Single-Index Models under the $L_2^2$ loss in the agnostic model. We give an efficient learning algorithm, achieving a constant factor approximation to the optimal loss, that succeeds und…

A Single Neuron Is Sufficient to Bypass Safety Alignment in Large Language Models

2026-05-08 · Hamid Kazemi, Atoosa Chegini, Maria Safi arxiv

Safety alignment in language models operates through two mechanistically distinct systems: refusal neurons that gate whether harmful knowledge is expressed, and concept neurons that encode the harmful knowledge itself. B…

Prompt Engineering

Trajectory Alignment: Understanding the Edge of Stability Phenomenon via Bifurcation Theory

2023-07-09 · NeurIPS 2023 11

Cohen et al. (2021) empirically study the evolution of the largest eigenvalue of the loss Hessian, also known as sharpness, along the gradient descent (GD) trajectory and observe the Edge of Stability (EoS) phenomenon. T…

Simplicity Bias via Global Convergence of Sharpness Minimization

2024-10-21 · Khashayar Gatmiry, Zhiyuan Li, Sashank J. Reddi, Stefanie Jegelka

The remarkable generalization ability of neural networks is usually attributed to the implicit bias of SGD, which often yields models with lower complexity using simpler (e.g. linear) and low-rank features. Recent works …

Enhancing Fine-Tuning Based Backdoor Defense with Sharpness-Aware Minimization

2023-04-24 · ICCV 2023 1 · Mingli Zhu, Shaokui Wei, Li Shen, Yanbo Fan 외

Backdoor defense, which aims to detect or mitigate the effect of malicious triggers introduced by attackers, is becoming increasingly critical for machine learning security and integrity. Fine-tuning based on benign data…

backdoor defense