Trustworthy Monte Carlo

Authors: Juha Harviainen, Mikko Koivisto, Petteri Kaski

NeurIPS 2022 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Theoretical Although this work is theoretical and leaves experimentation for future works, we include a proof-of-concept Mathematica [11] implementation for the example of Section 3.4 in the supplement.
Researcher Affiliation Academia Juha Harviainen University of Helsinki juha.harviainen@helsinki.fi Petteri Kaski Aalto University petteri.kaski@aalto.fi Mikko Koivisto University of Helsinki mikko.koivisto@helsinki.fi
Pseudocode Yes Algorithm V V1 Send the prover the problem instance and the points ξ1, ξ2, . . . , ξe. V2 Receive from the prover a proof, i.e., a claimed value yk of p(ξk) for each k = 1, 2, . . . , e. V3 Find the coefficients of p(x) = PD k=0 pkxk such that p(ξk) = yk for all k = 1, 2, . . . , e. V4 Draw a random point ξ0 F and evaluate p(ξ0) and p(ξ0). V5 If p(ξ0) = p(ξ0), then accept the proof and consume the values yk; otherwise reject the proof.
Open Source Code Yes Although this work is theoretical and leaves experimentation for future works, we include a proof-of-concept Mathematica [11] implementation for the example of Section 3.4 in the supplement.
Open Datasets No The paper describes theoretical methods and does not detail empirical experiments involving specific datasets, their public availability, or training procedures.
Dataset Splits No The paper is theoretical and does not describe empirical experiments, thus no training, validation, or test dataset splits are provided.
Hardware Specification No The paper is theoretical and does not describe empirical experiments, therefore no hardware specifications for running experiments are provided.
Software Dependencies Yes Wolfram Research, Inc. Mathematica, Version 13.0.0. Champaign, IL, 2021.
Experiment Setup No The paper is theoretical and does not describe empirical experiments or their setup, including hyperparameters or system-level training settings.