Notice: The reproducibility variables underlying each score are classified using an automated LLM-based pipeline, validated against a manually labeled dataset. LLM-based classification introduces uncertainty and potential bias; scores should be interpreted as estimates. Full accuracy metrics and methodology are described in [1].

A Unified Recipe for Deriving (Time-Uniform) PAC-Bayes Bounds

Authors: Ben Chugg, Hongjian Wang, Aaditya Ramdas

JMLR 2023 | Venue PDF | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Theoretical We present a unified framework for deriving PAC-Bayesian generalization bounds. Unlike most previous literature on this topic, our bounds are anytime-valid (i.e., time-uniform), meaning that they hold at all stopping times, not only for a fixed sample size. Our approach combines four tools in the following order: (a) nonnegative supermartingales or reverse submartingales, (b) the method of mixtures, (c) the Donsker-Varadhan formula (or other convex duality principles), and (d) Ville s inequality. Our main result is a PAC-Bayes theorem which holds for a wide class of discrete stochastic processes. We show how this result implies time-uniform versions of well-known classical PAC-Bayes bounds, such as those of Seeger, Mc Allester, Maurer, and Catoni, in addition to many recent bounds. We also present several novel bounds.
Researcher Affiliation Academia Ben Chugg EMAIL Hongjian Wang EMAIL Aaditya Ramdas EMAIL Departments of Statistics and Machine Learning Carnegie Mellon University
Pseudocode No The paper describes methods and derivations in prose and mathematical notation but does not include any explicitly labeled 'Pseudocode' or 'Algorithm' blocks or structured, code-like steps.
Open Source Code No The paper does not contain any explicit statements about making source code available, nor does it provide links to any code repositories.
Open Datasets No This is a theoretical paper focusing on deriving PAC-Bayesian bounds. It does not conduct empirical experiments on specific datasets, therefore, no datasets are used or made publicly available.
Dataset Splits No As this is a theoretical paper that does not perform empirical experiments using specific datasets, there is no mention of training/test/validation dataset splits.
Hardware Specification No The paper is theoretical and does not describe any experimental setup or hardware used for running experiments.
Software Dependencies No The paper is theoretical and focuses on mathematical derivations. It does not specify any software, libraries, or programming languages with version numbers required for reproduction.
Experiment Setup No This is a theoretical paper focused on deriving mathematical bounds. It does not include any experimental setup details, hyperparameters, or system-level training settings.