International Conference on Machine Learning (ICML)

Venue URL:

The Percentage of Empirical Papers Documenting Each Reproducibility Variable

Venue Year Papers
Repro. Score Reproducibility Score based on Gundersen et al. (2025)
Doc. Mean Global mean is the average score over the seven reproducibility variables for empirical research papers.
Doc. Median Global median is the median score over the seven reproducibility variables for empirical research papers.
Dataset Doc. Documentation mean is the average score over the Open Datasets and Dataset Splits reproducibility variables for empirical research papers.
Code Doc. Documentation mean is the average score over the Open Source Code reproducibility variables for empirical research papers.
Other Doc. Documentation mean is the average score over the Pseudocode, Hardware Specification, Software Dependencies, and Experiment Setup reproducibility variables for empirical research papers.
% Empirical Percentage of papers that are empirical research vs theoretical research
% Industry Percentage of empirical research papers with at least one author from Industry
Website
ICML 2025 3330 0.61 4.42 5.0 1.52 0.64 2.26 94.95% 36.69%
ICML 2024 2610 0.62 4.11 4.0 1.28 0.67 2.16 93.03% 39.99%
ICML 2023 1828 0.6 4.06 4.0 1.27 0.65 2.14 91.47% 43.84%
ICML 2022 1233 0.58 3.97 4.0 1.27 0.6 2.1 92.94% 42.84%
ICML 2021 1183 0.52 3.38 3.0 1.16 0.52 1.7 92.98% 45.09%
ICML 2020 1084 0.52 3.43 3.0 1.18 0.5 1.75 90.68% 44.05%
ICML 2019 773 0.49 3.23 3.0 1.15 0.46 1.62 92.24% 46.28%
ICML 2018 621 0.42 3.13 3.0 1.07 0.36 1.7 94.52% 40.2%
ICML 2017 434 0.39 3.15 3.0 1.13 0.26 1.76 92.17% 41.25%
ICML 2016 322 0.36 3.07 3.0 1.13 0.23 1.7 93.17% 33.0%
ICML 2015 270 0.37 3.24 3.0 1.2 0.23 1.82 94.07% 28.35%
ICML 2014 310 0.3 3.0 3.0 1.09 0.16 1.76 93.55% 27.59%