A unified recursion framework for stochastic variance-reduced estimation yields high-probability bounds and the first Õ(ε^{-3}) oracle complexity for stochastic optimization with expectation constraints.
Prodigy: An expeditiously adaptive parameter-free learner
8 Pith papers cite this work. Polarity classification is still indexing.
8
Pith papers citing it
citation-role summary
background 4
citation-polarity summary
roles
background 4polarities
background 4representative citing papers
Spectral clipping of leading singular values in gradient matrices stabilizes SGD for non-convex problems with heavy-tailed noise and achieves the optimal convergence rate O(K^{(2-2α)/(3α-2)}).
A 300K quadruplet dataset and UniDG foundation model enable reference- or text-driven defect generation across categories, outperforming few-shot baselines on anomaly detection tasks.
Empirical study of a fully synthetic data generation pipeline for text-based person retrieval that tests its use as a replacement or augmentation for real data across scenarios.
Optimizers like Adam reduce to steepest descent under particular norms, opening a design space of norm assignments tailored to layer roles.
SGD is reformulated via a master equation from discrete updates, producing a discrete Fokker-Planck equation that predicts non-stationary variance growth proportional to learning rate in flat Hessian directions.
ImgEdit supplies 1.2 million curated edit pairs and a three-part benchmark that let a VLM-based model outperform prior open-source editors on adherence, quality, and detail preservation.
A two-stage method predicts an intermediate Canny map for structure then renders the image conditioned on appearance and structure, paired with a 100k text-aware dataset, to improve detail preservation in subject-driven generation.
citing papers explorer
-
Unified High-Probability Analysis of Stochastic Variance-Reduced Estimation
A unified recursion framework for stochastic variance-reduced estimation yields high-probability bounds and the first Õ(ε^{-3}) oracle complexity for stochastic optimization with expectation constraints.
-
Gradient Clipping Beyond Vector Norms: A Spectral Approach for Matrix-Valued Parameters
Spectral clipping of leading singular values in gradient matrices stabilizes SGD for non-convex problems with heavy-tailed noise and achieves the optimal convergence rate O(K^{(2-2α)/(3α-2)}).
-
Large-Scale Universal Defect Generation: Foundation Models and Datasets
A 300K quadruplet dataset and UniDG foundation model enable reference- or text-driven defect generation across categories, outperforming few-shot baselines on anomaly detection tasks.
-
An Empirical Study of Validating Synthetic Data for Text-Based Person Retrieval
Empirical study of a fully synthetic data generation pipeline for text-based person retrieval that tests its use as a replacement or augmentation for real data across scenarios.
-
Old Optimizer, New Norm: An Anthology
Optimizers like Adam reduce to steepest descent under particular norms, opening a design space of norm assignments tailored to layer roles.
-
Why SGD is not Brownian Motion: A New Perspective on Stochastic Dynamics
SGD is reformulated via a master equation from discrete updates, producing a discrete Fokker-Planck equation that predicts non-stationary variance growth proportional to learning rate in flat Hessian directions.
-
ImgEdit: A Unified Image Editing Dataset and Benchmark
ImgEdit supplies 1.2 million curated edit pairs and a three-part benchmark that let a VLM-based model outperform prior open-source editors on adherence, quality, and detail preservation.
-
Decomposing Subject-Driven Image Generation via Intermediate Structural Prediction
A two-stage method predicts an intermediate Canny map for structure then renders the image conditioned on appearance and structure, paired with a 100k text-aware dataset, to improve detail preservation in subject-driven generation.