# ICML

> Source: https://aiwiki.ai/wiki/icml
> Updated: 2026-06-21
> Categories: AI Events, Machine Learning
> From AI Wiki (https://aiwiki.ai), a free encyclopedia of artificial intelligence. Quote with attribution.

**ICML**, the **International Conference on Machine Learning**, is one of the three top-tier academic conferences in [machine learning](/wiki/machine_learning), alongside [NeurIPS](/wiki/neurips) and [ICLR](/wiki/iclr), and is organized annually by the [International Machine Learning Society](/wiki/imls) (IMLS). It grew out of an informal workshop series first held at [Carnegie Mellon University](/wiki/cmu) in 1980 and has since become one of the largest gatherings in artificial intelligence: ICML 2025 in Vancouver drew 12,107 paper submissions and accepted 3,260 of them, a 26.9 percent acceptance rate [1][2][9]. The conference publishes its proceedings open access through the Proceedings of Machine Learning Research (PMLR) [3].

ICML covers the full breadth of machine learning research, including supervised, unsupervised, and reinforcement learning, statistical learning theory, optimization for ML, probabilistic models, kernel methods, [deep learning](/wiki/deep_learning), representation learning, and the interfaces with natural language processing, computer vision, robotics, and computational biology. Proceedings have been published open access since 2010 through the Proceedings of Machine Learning Research (PMLR) [3]. The official ICML 2026 Call for Papers describes the venue as a home for "original and rigorous research of significant interest to the machine learning community" [26].

## History

### Origins as a workshop (1980 to 1987)

The lineage of ICML traces back to the first **International Workshop on Machine Learning** (IWML), held in July 1980 at [Carnegie Mellon University](/wiki/cmu) in Pittsburgh. The 1980 workshop was organized by [Ryszard S. Michalski](/wiki/ryszard_michalski), [Jaime Carbonell](/wiki/jaime_carbonell), and [Tom M. Mitchell](/wiki/tom_mitchell), three figures who would go on to edit the foundational *Machine Learning: An Artificial Intelligence Approach* book series that helped define machine learning as a coherent subfield distinct from broader artificial intelligence [1][4]. The workshop format continued through the 1980s, with subsequent meetings in 1983 (Monticello, Illinois), 1985 (Skytop, Pennsylvania), and 1987 (Irvine, California). These were intimate gatherings of a few dozen to a few hundred attendees, with proceedings often published as edited volumes.

### Transition to a conference (1988 onward)

ICML adopted the conference name and numbering scheme starting with the **5th International Conference on Machine Learning** in 1988, held in Ann Arbor, Michigan. The numbering reflects continuity with the four prior workshops [1]. The 1993 edition at the University of Massachusetts Amherst is sometimes cited as the point at which ICML solidified its modern conference format with refereed proceedings, although the conference designation had begun in 1988 [1].

### Founding of the IMLS (2001 to 2002)

For most of its early history, ICML was run informally, with each year's host institution managing logistics independently. At the **ICML 2001 business meeting** in Williamstown, Massachusetts, attendees approved the creation of a democratic governance structure, and the **[International Machine Learning Society](/wiki/imls)** (IMLS) was established as a non-profit organization in 2001, with formal incorporation completed shortly thereafter [5][6]. The IMLS now governs ICML, manages the conference finances, and oversees affiliated publishing efforts. The IMLS is distinct from the Neural Information Processing Systems Foundation, which separately governs [NeurIPS](/wiki/neurips).

### Growth in the 2010s and 2020s

The rise of [deep learning](/wiki/deep_learning) after 2012 transformed ICML from a moderately sized academic conference into one of the largest events in computer science. Submissions roughly doubled every few years through the 2010s: the conference received 1,037 submissions in 2015 and 12,107 in 2025, more than an elevenfold increase in a decade [9][10]. The conference also expanded to include extensive workshop and tutorial tracks, and by the mid-2020s ICML had become a major commercial as well as academic event, with sponsorship from industry research labs playing a substantial role [2].

## Format

A typical modern ICML runs for five to seven days and has a layered structure that has evolved gradually since the 1990s.

* **Tutorials.** Half-day or full-day sessions, usually held on the first day, that survey emerging subfields or methods.
* **Main conference.** The core technical program runs for three to four days. Accepted papers are presented either as oral talks (a small fraction, typically 1 to 2 percent) or as posters at large evening sessions. At ICML 2025, oral presentations accounted for roughly the top 1 percent of submissions (108 papers), spotlight posters for 2.6 percent (313 papers), and standard posters for the remainder of the 3,260 accepted papers [9][10].
* **Keynote and invited talks.** Senior researchers from academia and industry deliver plenary addresses on themes such as foundation models, learning theory, or AI safety.
* **Workshops.** One- or two-day topical workshops, held on the days bracketing the main conference, focus on niche or emerging areas. Workshop papers are reviewed separately from the main track and are generally non-archival.
* **Position track.** Introduced for ICML 2024, the position paper track invites submissions that argue for a particular stance on issues facing the machine learning community, including research priorities, ethics, governance, or methodology. Position submissions undergo double-blind [peer review](/wiki/peer_review) under criteria that emphasize whether the paper presents a compelling, well-supported perspective rather than novel empirical results [7]. The track was retained in 2025 and 2026.
* **Volunteer program.** ICML offers a long-running volunteer program that provides registration discounts to graduate students who help with on-site logistics.
* **Affinity events.** Groups such as Women in Machine Learning, LatinX in AI, Black in AI, and Queer in AI typically host affinity events alongside ICML.

Reviewing for the main track has been double-blind since the early 2010s. Different years have used different review platforms; recent editions have made use of [OpenReview](/wiki/openreview) for portions of the review cycle, while submission and reviewer assignment have at times been managed through CMT (Microsoft's Conference Management Toolkit).

## Locations and editions

The table below lists ICML editions from the 1980 founding workshop through the most recent meetings. Conference numbering follows the IMLS convention of counting from the 1980 IWML as the first edition [1].

| Year | Edition | Location | Country |
| --- | --- | --- | --- |
| 1980 | 1st (IWML) | Pittsburgh ([CMU](/wiki/cmu)) | United States |
| 1983 | 2nd | Monticello, Illinois | United States |
| 1985 | 3rd | Skytop, Pennsylvania | United States |
| 1987 | 4th | Irvine, California | United States |
| 1988 | 5th | Ann Arbor, Michigan | United States |
| 1992 | 9th | Aberdeen | United Kingdom |
| 1993 | 10th | Amherst, Massachusetts | United States |
| 1996 | 13th | Bari | Italy |
| 1999 | 16th | Bled | Slovenia |
| 2001 | 18th | Williamstown, Massachusetts | United States |
| 2002 | 19th | Sydney | Australia |
| 2004 | 21st | Banff, Alberta | Canada |
| 2005 | 22nd | Bonn | Germany |
| 2008 | 25th | Helsinki | Finland |
| 2009 | 26th | Montreal | Canada |
| 2010 | 27th | Haifa | Israel |
| 2011 | 28th | Bellevue, Washington | United States |
| 2012 | 29th | Edinburgh | United Kingdom |
| 2013 | 30th | Atlanta, Georgia | United States |
| 2014 | 31st | Beijing | China |
| 2015 | 32nd | Lille | France |
| 2016 | 33rd | New York City | United States |
| 2017 | 34th | Sydney | Australia |
| 2018 | 35th | Stockholm | Sweden |
| 2019 | 36th | Long Beach, California | United States |
| 2020 | 37th | Vienna (virtual due to COVID-19) | Austria |
| 2021 | 38th | Vienna (virtual) | Austria |
| 2022 | 39th | Baltimore, Maryland | United States |
| 2023 | 40th | Honolulu, Hawaii | United States |
| 2024 | 41st | Vienna | Austria |
| 2025 | 42nd | Vancouver | Canada |
| 2026 | 43rd | Seoul | South Korea |

ICML 2014 in Beijing was the first time the conference was held in mainland China and marked an important step in broadening its geographic reach. The 2020 and 2021 editions were forced online by the COVID-19 pandemic. Both years are still nominally associated with Vienna, Austria, where they had been scheduled to take place; the conference returned to Vienna for an in-person meeting in 2024 [8].

### When and where is ICML 2026?

ICML 2026 is the 43rd edition and is scheduled for July 6 to 11, 2026, at the COEX Convention and Exhibition Center in Seoul, South Korea, the first time the conference is held in South Korea [26][27]. The main paper submission deadline was January 28, 2026, with abstracts due January 23 [26]. In a May 2026 registration update, the organizers reported that the conference was "very close to capacity" and that general registration for the main conference was "expected to sell out," with a cap set "marginally below the legal limit for the venue" [28]. The position-paper track introduced in 2024 continued for the 2026 edition [26].

### What were the ICML 2025 statistics?

ICML 2025, the 42nd edition, was held July 13 to 19, 2025, at the Vancouver Convention Centre in Vancouver, Canada [27]. It received 12,107 submissions and accepted 3,260 papers, a 26.9 percent acceptance rate, the largest submission count in the conference's history at that point [9][10]. Of the accepted papers, 108 (about 1 percent of all submissions) were selected as oral presentations and 313 (2.6 percent) as spotlight posters [9][10].

### Submission volumes and acceptance rates

From the 1990s through the early 2010s, ICML typically saw a few hundred to roughly a thousand submissions per year, with [acceptance rates](/wiki/acceptance_rate) hovering between 20 and 30 percent. The deep-learning era pushed those numbers sharply upward.

| Year | Submissions | Accepted | Acceptance rate |
| --- | --- | --- | --- |
| 2015 | 1,037 | 270 | 26 percent |
| 2018 | 2,473 | 621 | 25 percent |
| 2020 | 4,990 | 1,088 | 21.8 percent |
| 2022 | 5,630 | 1,235 | 21.9 percent |
| 2024 | 9,473 | 2,609 | 27.5 percent |
| 2025 | 12,107 | 3,260 | 26.9 percent |

These figures are sourced from official ICML statistics pages and the IMLS fact sheets where available [9][10][11]. The 2024 acceptance rate, around 27.5 percent, was unusually high relative to recent precedent, in part because that year saw an enlarged program committee and a deliberate increase in the number of accepted papers to absorb the growing submission load [11]. The 2024 program included 144 oral presentations and 191 spotlights among its accepted papers [11].

## Awards

ICML presents two principal classes of awards each year: paper awards for outstanding work submitted to the current conference, and a Test of Time Award recognizing influential papers from a decade earlier. The terminology has shifted over time. Through 2018 ICML primarily used the label "Best Paper Award"; since 2019 the conference has often used "Outstanding Paper Award" for the same recognition, with Best Paper Award occasionally retained for top selections among the outstanding set. The two terms are used interchangeably in much community discussion.

### Best Paper and Outstanding Paper awards

The table below summarizes selected paper award recipients across recent years. It is not exhaustive: most years select multiple papers and additional honorable mentions or runners-up.

| Year | Paper | Authors | Award |
| --- | --- | --- | --- |
| 2017 | Understanding Black-box Predictions via Influence Functions | Pang Wei Koh, Percy Liang | [Best Paper](/wiki/best_paper_award) [12] |
| 2018 | Obfuscated Gradients Give a False Sense of Security | Anish Athalye, Nicholas Carlini, David Wagner | Best Paper [13] |
| 2018 | Delayed Impact of Fair Machine Learning | Lydia Liu, Sarah Dean, Esther Rolf, Max Simchowitz, Moritz Hardt | Best Paper [13] |
| 2019 | Challenging Common Assumptions in the Unsupervised Learning of Disentangled Representations | Francesco Locatello, Stefan Bauer, Mario Lucic, Gunnar Ratsch, Sylvain Gelly, Bernhard Schoelkopf, Olivier Bachem | Best Paper [14] |
| 2019 | Rates of Convergence for Sparse Variational Gaussian Process Regression | David Burt, Carl Rasmussen, Mark van der Wilk | Best Paper [14] |
| 2020 | On Learning Sets of Symmetric Elements | Haggai Maron, Or Litany, Gal Chechik, Ethan Fetaya | Outstanding Paper [15] |
| 2020 | Tuning-free Plug-and-Play Proximal Algorithm for Inverse Imaging Problems | Kaixuan Wei, Angelica Aviles-Rivero, Jingwei Liang, Ying Fu, Carola-Bibiane Schoenlieb, Hua Huang | Outstanding Paper [15] |
| 2021 | Unbiased Gradient Estimation in Unrolled Computation Graphs with Persistent Evolution Strategies | Paul Vicol, Luke Metz, Jascha Sohl-Dickstein | Outstanding Paper [16] |
| 2022 | Stable Conformal Prediction Sets | Eugene Ndiaye | Outstanding Paper [17] |
| 2022 | Causal Conceptions of Fairness and their Consequences | Hamed Nilforoshan, Johann Gaebler, Ravi Shroff, Sharad Goel | Outstanding Paper [17] |
| 2022 | G-Mixup: Graph Data Augmentation for Graph Classification | Xiaotian Han, Zhimeng Jiang, Ninghao Liu, Xia Hu | Outstanding Paper [17] |
| 2023 | Learning-Rate-Free Learning by D-Adaptation | Aaron Defazio, Konstantin Mishchenko | Outstanding Paper [18] |
| 2023 | A Watermark for Large Language Models | John Kirchenbauer, Jonas Geiping, Yuxin Wen, Jonathan Katz, Ian Miers, Tom Goldstein | Outstanding Paper [18] |
| 2024 | Stealing Part of a Production Language Model | Nicholas Carlini, Daniel Paleka, and others | Best Paper [19] |
| 2024 | Genie: Generative Interactive Environments | Jake Bruce, Michael Dennis, Ashley Edwards, and others | Best Paper [19] |
| 2024 | Discrete Diffusion Modeling by Estimating the Ratios of the Data Distribution | Aaron Lou, Chenlin Meng, Stefano Ermon | Best Paper [19] |
| 2024 | Position: Considerations for Differentially Private Learning with Large-Scale Public Pretraining | Florian Tramer, Gautam Kamath, Nicholas Carlini | Best Paper (position track) [19] |
| 2024 | Position: Measure Dataset Diversity, Don't Just Claim It | Dora Zhao, Jerone Andrews, Orestis Papakyriakopoulos, Alice Xiang | Best Paper (position track) [19] |
| 2024 | Debating with More Persuasive LLMs Leads to More Truthful Answers | Akbir Khan, John Hughes, and others | Best Paper [19] |
| 2024 | VideoPoet: A Large Language Model for Zero-Shot Video Generation | Dan Kondratyuk, Lijun Yu, and others | Best Paper [19] |
| 2025 | CollabLLM: From Passive Responders to Active Collaborators | Shirley Wu, Michel Galley, Baolin Peng, and others | Outstanding Paper [20] |
| 2025 | Train for the Worst, Plan for the Best | Jaeyeon Kim, Kulin Shah, Vasilis Kontonis, Sham Kakade, Sitan Chen | Outstanding Paper [20] |
| 2025 | Conformal Prediction as Bayesian Quadrature | Jake Snell, Thomas Griffiths | Outstanding Paper [20] |
| 2025 | The Value of Prediction in Identifying the Worst-Off | Unai Fischer Abaigar, Christoph Kern, Juan Perdomo | Outstanding Paper [20] |

The 2024 awards drew particular attention because three of the recognized works fell under the new position-paper track (the differentially private pretraining and dataset-diversity papers being the most discussed), which gave the inaugural year of that track an immediate seal of legitimacy [7][19].

### [Test of Time](/wiki/test_of_time) Award

ICML introduced its Test of Time Award in 2009 with retrospective recognition of papers presented a decade earlier. The award was put on hold for several years and then restarted in 2017 in its current form, recognizing one paper from the conference held ten years prior that has had outstanding lasting impact. Recent recipients include the following.

| Year given | Paper | Year published | Authors |
| --- | --- | --- | --- |
| 2017 | Combining Online and Offline Knowledge in UCT | 2007 | Sylvain Gelly, David Silver |
| 2018 | A Unified Architecture for Natural Language Processing: Deep Neural Networks with Multitask Learning | 2008 | Ronan Collobert, Jason Weston |
| 2019 | Online Dictionary Learning for Sparse Coding | 2009 | Julien Mairal, Francis Bach, Jean Ponce, Guillermo Sapiro |
| 2020 | Gaussian Process Optimization in the Bandit Setting: No Regret and Experimental Design | 2010 | Niranjan Srinivas, Andreas Krause, Sham Kakade, Matthias Seeger |
| 2021 | Bayesian Learning via Stochastic Gradient Langevin Dynamics | 2011 | Max Welling, Yee Whye Teh |
| 2022 | Poisoning Attacks Against Support Vector Machines | 2012 | Battista Biggio, Blaine Nelson, Pavel Laskov |
| 2023 | Learning Fair Representations | 2013 | Rich Zemel, Yu Wu, Kevin Swersky, Toni Pitassi, Cynthia Dwork |
| 2024 | DeCAF: A Deep Convolutional Activation Feature for Generic Visual Recognition | 2014 | Jeffrey Donahue, Yangqing Jia, Oriol Vinyals, Judy Hoffman, Ning Zhang, Eric Tzeng, Trevor Darrell |
| 2025 | Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift | 2015 | Sergey Ioffe, Christian Szegedy |

The selection of "Combining Online and Offline Knowledge in UCT" by Gelly and Silver in 2017 was widely viewed as fitting given the role that Monte Carlo tree search later played in [deep reinforcement learning](/wiki/deep_learning) systems including AlphaGo. The 2025 recognition of Batch Normalization underlines how thoroughly that technique reshaped deep network training [21].

## Notable papers

Many influential works in modern machine learning first appeared at ICML. Selected papers that genuinely originated at ICML and have proven highly influential include the following.

* **Stochastic Variational Inference** (Hoffman, Blei, Wang, Paisley) appeared in JMLR Volume 14 in 2013. It generalized variational inference to large datasets and remains a workhorse method for scalable Bayesian inference [22].
* **Trust Region Policy Optimization** (Schulman, Levine, Abbeel, Jordan, Moritz) was published at ICML 2015. TRPO provided a principled approach to policy gradient methods with monotonic improvement guarantees and was the direct precursor to Proximal Policy Optimization (PPO) [23].
* **Asynchronous Methods for Deep Reinforcement Learning** (Mnih, Badia, Mirza, Graves, Lillicrap, Harley, Silver, Kavukcuoglu) was published at ICML 2016. Better known as the A3C paper, it introduced the asynchronous advantage actor-critic framework that dominated [deep reinforcement learning](/wiki/reinforcement_learning) research for several years [24].
* **Batch Normalization** (Ioffe, Szegedy) was published at ICML 2015 and won the 2025 Test of Time Award; the paper introduced one of the most widely deployed normalization techniques in [deep learning](/wiki/deep_learning) [21].
* **Understanding Black-box Predictions via Influence Functions** (Koh, Liang) won the 2017 Best Paper Award and helped revive influence functions as a tool for interpretability and data attribution [12].
* **Challenging Common Assumptions in the Unsupervised Learning of Disentangled Representations** (Locatello and colleagues) won the 2019 Best Paper Award and showed that purely unsupervised disentanglement is impossible without inductive biases [14].
* **Stealing Part of a Production Language Model** (Carlini and colleagues) won a 2024 Best Paper Award and was the first published model-extraction attack to recover non-trivial information from black-box production large language models [19].
* **Genie: Generative Interactive Environments** (Bruce, Dennis, Edwards, Parker-Holder, and colleagues) won a 2024 Best Paper Award for demonstrating the first generative interactive environment trained from unlabeled internet videos [19].

It is worth noting which famous papers are commonly but incorrectly associated with ICML. The original AlexNet paper (Krizhevsky, Sutskever, Hinton, 2012) was at NeurIPS, not ICML. The first Generative Adversarial Networks paper (Goodfellow and colleagues, 2014) was at NeurIPS. The Adam optimizer paper (Kingma and Ba, 2015) was at ICLR. Word2vec (Mikolov and colleagues, 2013) appeared at ICLR and NeurIPS. The neural attention paper for machine translation (Bahdanau, Cho, Bengio, 2015) was at ICLR. The original AlphaGo paper appeared in *Nature*, not at any of the top ML conferences. The official PMLR proceedings and DBLP listings are authoritative for venue attribution.

## Comparison with other ML conferences

ICML is one of three flagship conferences for machine learning research. The table below compares it with [NeurIPS](/wiki/neurips), [ICLR](/wiki/iclr), and other adjacent venues. Numbers are approximate and reflect recent multi-year averages.

| Conference | Founded | Governance | Scope | Typical submissions | Typical [acceptance rate](/wiki/acceptance_rate) |
| --- | --- | --- | --- | --- | --- |
| [ICML](/wiki/icml) | 1980 (workshop), 1988 (conference) | [IMLS](/wiki/imls) | All of [machine learning](/wiki/machine_learning) | 9,000 to 12,000 | 22 to 28 percent |
| [NeurIPS](/wiki/neurips) | 1987 | NeurIPS Foundation | ML, neuroscience, statistics, applied probability | 12,000 to 17,000 | 23 to 27 percent |
| [ICLR](/wiki/iclr) | 2013 | Open Review (community) | Representation learning, [deep learning](/wiki/deep_learning) | 7,000 to 12,000 | 27 to 32 percent |
| [AAAI](/wiki/aaai) | 1980 | AAAI | Broad AI including ML, KR, planning | 8,000 to 14,000 | 20 to 25 percent |
| [IJCAI](/wiki/ijcai) | 1969 | IJCAI | Broad AI | 4,000 to 7,000 | 13 to 17 percent |
| [CVPR](/wiki/cvpr) | 1985 (formal) | IEEE / CVF | Computer vision | 9,000 to 13,000 | 22 to 28 percent |
| [ECCV](/wiki/eccv) | 1990 | ECVA | Computer vision | 5,000 to 9,000 | 25 to 30 percent |
| [ICCV](/wiki/iccv) | 1987 | IEEE / CVF | Computer vision | 7,000 to 12,000 | 23 to 27 percent |
| [ACL](/wiki/acl) | 1962 | ACL | Computational linguistics | 4,000 to 7,000 | 20 to 25 percent |
| [EMNLP](/wiki/emnlp) | 1996 | ACL SIGDAT | NLP, empirical methods | 4,000 to 7,000 | 20 to 25 percent |
| [NAACL](/wiki/naacl) | 2000 | ACL | NLP, North American chapter | 3,000 to 5,000 | 22 to 28 percent |

### How does ICML differ from NeurIPS and ICLR?

A common point of confusion is the relationship between ICML and [NeurIPS](/wiki/neurips). Both are top-tier general machine learning conferences with substantial overlap in scope and authorship, but they are governed by entirely separate non-profit bodies. ICML is run by the [IMLS](/wiki/imls), founded in 2001 to 2002, while NeurIPS is run by the Neural Information Processing Systems Foundation, which traces its origin to 1987. [ICLR](/wiki/iclr), first held in 2013, has historically focused more narrowly on representation learning and pioneered fully open peer review through [OpenReview](/wiki/openreview). In practice, ICML is typically held in July, NeurIPS in December, and ICLR in the spring, so the three flagship venues space the machine learning publishing calendar across the year.

## Recent trends

Several trends have shaped ICML in the last several years.

* **Submission volumes and reviewer load.** Submissions roughly tripled between 2018 and 2024 and continued to climb in 2025, reaching 12,107. Maintaining review quality at this scale has become a perennial concern, with each program committee struggling to recruit enough qualified reviewers and area chairs [9].
* **Surge in LLM and foundation model submissions.** From 2022 onward, large language model and foundation-model topics came to dominate accepted papers, mirroring trends across [NeurIPS](/wiki/neurips) and [ICLR](/wiki/iclr). Subtopics including alignment, evaluation, fine-tuning, retrieval-augmented generation, agents, and safety became major program tracks.
* **Position track.** The 2024 introduction of a position-paper track was a deliberate attempt to give the community a venue for non-empirical arguments. The recognition of multiple position papers with Best Paper Awards in 2024 signaled that the IMLS intended the track to have substantive standing [7].
* **Reproducibility and code release.** ICML expanded its reproducibility checklist initiative through the late 2010s, asking authors to disclose dataset preparation, training compute, and ablation details.
* **Reviewing platforms.** ICML has moved between Microsoft CMT and [OpenReview](/wiki/openreview) for various reviewing tasks, and recent years have experimented with reviewer rewards and stronger desk-rejection criteria.
* **Industrial sponsorship.** Sponsorship from large industrial AI labs has become a substantial fraction of conference revenue, and recruiting events now rival the technical program in attendance for some attendees.
* **Capacity limits.** As demand has outstripped venue space, ICML has begun capping registration. Ahead of the 2026 Seoul edition, organizers warned in May 2026 that the conference was "very close to capacity" and that general registration was "expected to sell out" [28].
* **Hybrid and virtual modes.** After the fully virtual editions in 2020 and 2021, the conference settled into an in-person primary mode with virtual streaming and a virtual poster site for accessibility.

## Criticisms

Like other top machine learning conferences, ICML has been the subject of recurring criticism from the community.

* **Acceptance-rate inflation and submission-volume strain.** As submissions surge past the program committee's capacity to review carefully, reviewers see an increasing number of papers each, often with insufficient time. Critics argue that this leads to higher noise in accept and reject decisions.
* **Reviewer quality concerns.** Reports of perfunctory reviews, factually wrong reviews, and reviewers who admit to not reading the paper appear regularly during ICML rebuttal seasons. Recent program chairs have responded with reviewer training, mandatory reviewer feedback, and in some cases reviewer disqualification policies.
* **Position-paper debates.** The 2024 position-paper track sparked controversy over which kinds of arguments should be eligible for paper awards. Critics argued that judging position papers under blind peer review favors well-known authors with established credibility on contested topics.
* **Double-submission and deduplication policies.** ICML, [NeurIPS](/wiki/neurips), and [ICLR](/wiki/iclr) have overlapping deadlines and similar scope, leading to repeated debates about whether papers rejected at one venue should be permitted to be lightly revised and resubmitted to the next.
* **Fairness, ethics, and safety review.** ICML introduced an ethics review process for papers raising potential societal harms in the late 2010s, and subsequent years have seen disagreement over how stringently it should be applied.
* **Award controversies.** Specific award decisions have occasionally drawn pushback, including a 2023 public discussion in which a researcher argued that citation practices around an award-winning paper were uneven.
* **Travel costs and accessibility.** As venues rotate among expensive convention centers and registration fees climb, attendees from less-resourced institutions report difficulty attending. The IMLS has expanded subsidies and travel-grant programs in response.

## See also

* [NeurIPS](/wiki/neurips)
* [ICLR](/wiki/iclr)
* [AAAI](/wiki/aaai)
* [IJCAI](/wiki/ijcai)
* [CVPR](/wiki/cvpr)
* [ACL](/wiki/acl)
* [EMNLP](/wiki/emnlp)
* [NAACL](/wiki/naacl)
* [Machine learning](/wiki/machine_learning)
* [Deep learning](/wiki/deep_learning)
* [Reinforcement learning](/wiki/reinforcement_learning)
* [International Machine Learning Society](/wiki/imls)
* [OpenReview](/wiki/openreview)
* [Peer review](/wiki/peer_review)
* [Test of Time award](/wiki/test_of_time)
* [Best paper award](/wiki/best_paper_award)

## References

1. Wikipedia. "International Conference on Machine Learning." https://en.wikipedia.org/wiki/International_Conference_on_Machine_Learning
2. International Conference on Machine Learning. "About." https://icml.cc/About
3. Proceedings of Machine Learning Research. https://proceedings.mlr.press/
4. Machine Learning and Inference Laboratory, George Mason University. "Photos from Conferences." https://www.mli.gmu.edu/michalski/confphotos.html
5. International Machine Learning Society. "About." https://machinelearning.org/
6. Mooney, Raymond, and Dietterich, Thomas. "Starting an International Machine Learning Society (IMLS)." https://www.cs.utexas.edu/~ml/election/imls.html
7. ICML 2024 Call for Position Papers. https://icml.cc/Conferences/2024/CallForPositionPapers
8. ICML 2024 Conference. https://icml.cc/Conferences/2024
9. ICML 2024 Fact Sheet. https://media.icml.cc/Conferences/ICML2024/ICML2024_Fact_Sheet.pdf
10. Paper Copilot ICML Statistics. https://papercopilot.com/statistics/icml-statistics/
11. ICML 2024 Conference page. https://icml.cc/Conferences/2024
12. Koh, Pang Wei, and Liang, Percy. "Understanding Black-box Predictions via Influence Functions." ICML 2017. https://proceedings.mlr.press/v70/koh17a.html
13. ICML 2018 Awards. https://icml.cc/Conferences/2018/Awards
14. ICML 2019 Awards. https://icml.cc/virtual/2019/awards_detail
15. ICML 2020 Awards. https://icml.cc/Conferences/2020/Awards
16. ICML 2021 Awards. https://icml.cc/virtual/2021/awards_detail
17. ICML 2022 Awards. https://icml.cc/virtual/2022/awards_detail
18. ICML 2023 Outstanding Paper Awards. https://icml.cc/Conferences/2023/Awards
19. ICML 2024 Awards. https://icml.cc/virtual/2024/awards_detail
20. ICML 2025 Awards. https://icml.cc/virtual/2025/awards_detail
21. Ioffe, Sergey, and Szegedy, Christian. "Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift." ICML 2015. https://proceedings.mlr.press/v37/ioffe15.html
22. Hoffman, Matthew D., Blei, David M., Wang, Chong, and Paisley, John. "Stochastic Variational Inference." Journal of Machine Learning Research, Vol. 14, 2013. https://jmlr.org/papers/v14/hoffman13a.html
23. Schulman, John, Levine, Sergey, Abbeel, Pieter, Jordan, Michael, and Moritz, Philipp. "Trust Region Policy Optimization." ICML 2015. https://proceedings.mlr.press/v37/schulman15.html
24. Mnih, Volodymyr, Badia, Adria, Mirza, Mehdi, Graves, Alex, Lillicrap, Timothy, Harley, Tim, Silver, David, and Kavukcuoglu, Koray. "Asynchronous Methods for Deep Reinforcement Learning." ICML 2016. https://proceedings.mlr.press/v48/mniha16.html
25. DBLP entry for ICML. https://dblp.org/db/conf/icml/index.html
26. ICML 2026 Call for Papers. https://icml.cc/Conferences/2026/CallForPapers
27. ICML Future Meetings. https://icml.cc/Conferences/FutureMeetings
28. ICML Blog. "ICML 2026 Registration Update: Capacity Limits and What Attendees Need to Know." May 24, 2026. https://blog.icml.cc/2026/05/24/icml-2026-registration-update-capacity-limits-and-what-attendees-need-to-know/