zeroth order - Robuta Search

https://openreview.net/forum?id=fPwWfoyxL1 Riemannian Accelerated Zeroth-order Algorithm: Improved Robustness and Lower Query Complexity |... Optimization problems with access to only zeroth-order information of the objective function on Riemannian manifolds arise in various applications, spanning... zeroth order riemannian accelerated algorithm https://openreview.net/forum?id=BJe-DsC5Fm signSGD via Zeroth-Order Oracle | OpenReview We design and analyze a new zeroth-order stochastic optimization algorithm, ZO-signSGD, and demonstrate its connection and application to black-box adversarial... zeroth order via oracle openreview https://openreview.net/forum?id=ryxz8CVYDH Learning to Learn by Zeroth-Order Oracle | OpenReview Novel variant of learning to learn framework for zeroth-order optimization that learns both the update rule and the Gaussian sampling rule. learning to learn zeroth order oracle openreview https://openreview.net/forum?id=IzqZbNMZ0M Private Zeroth-Order Nonsmooth Nonconvex Optimization | OpenReview We introduce a new zeroth-order algorithm for private stochastic optimization on nonconvex and nonsmooth objectives. Given a dataset of size $M$, our algorithm... zeroth order private nonconvex optimization openreview https://openreview.net/forum?id=2ZNPedOfwB Zeroth-Order Hard-Thresholding: Gradient Error vs. Expansivity | OpenReview A Zeroth-Order Hard-Thresholding algorithm to reduce the dependence on the dimensionality. zeroth order hard thresholding gradient error https://openreview.net/forum?id=yrqn9rQO2YT Instance-Dependent Bounds for Zeroth-order Lipschitz Optimization with Error Certificates |... We characterize the sample complexity of certified Lipschitz optimization zeroth order instance dependent bounds https://arxiv.org/abs/2306.05655v1 [2306.05655v1] Communication-Efficient Zeroth-Order Distributed Online Optimization: Algorithm,... Abstract page for arXiv paper 2306.05655v1: Communication-Efficient Zeroth-Order Distributed Online Optimization: Algorithm, Theory, and Applications zeroth order online optimization 2306 communication efficient https://arxiv.org/abs/2304.01797 [2304.01797] Safe Zeroth-Order Optimization Using Linear Programs Abstract page for arXiv paper 2304.01797: Safe Zeroth-Order Optimization Using Linear Programs zeroth order 2304 01797 safe optimization https://openreview.net/forum?id=peyLy5ek4w&referrer=%5Bthe%20profile%20of%20Leonardo%20Lucio%20Custode%5D(%2Fprofile%3Fid%3D~Leonardo_Lucio_Custode1) Zeroth-Order Adaptive Neuron Alignment Based Pruning without Re-Training | OpenReview Network pruning focuses on computational techniques that aim to reduce a given model's computational cost by removing a subset of its parameters while having... zeroth order re training adaptive neuron alignment https://openreview.net/forum?id=qBWhjsNPEY DeepZero: Scaling Up Zeroth-Order Optimization for Deep Model Training | OpenReview Zeroth-order (ZO) optimization has become a popular technique for solving machine learning (ML) problems when first-order (FO) information is difficult or... scaling up zeroth order model training https://openreview.net/forum?id=THPjMr2r0S Revisiting Zeroth-Order Optimization for Memory-Efficient LLM Fine-Tuning: A Benchmark | OpenReview In the evolving landscape of natural language processing (NLP), fine-tuning pre-trained Large Language Models (LLMs) with first-order (FO) optimizers like SGD... llm fine tuning https://openreview.net/forum?id=W9G_ImpHlQd How to Robustify Black-Box ML Models? A Zeroth-Order Optimization Perspective | OpenReview The lack of adversarial robustness has been recognized as an important issue for state-of-the-art machine learning (ML) models, e.g., deep neural networks... https://openreview.net/forum?id=v5WnhGQmWl&referrer=%5Bthe%20profile%20of%20Qingcai%20Chen%5D(%2Fprofile%3Fid%3D~Qingcai_Chen2) ZO-AdaMU Optimizer: Adapting Perturbation by the Momentum and Uncertainty in Zeroth-Order... Lowering the memory requirement in full-parameter training on large models has become a hot research area. MeZO fine-tunes the large language models (LLMs) by...