\n 
\n {`${'CLeaR ' + year}`}\n
\n | Name | \nDate | \n
|---|---|
| {name} | \n{date} | \n
\n {position}\n {persons.map(({ name, href, address }) => (\n \n ))}\n | \n
Thank you for agreeing to review CLearR 2022! Your assessment is vital to creating a high quality program. This page provides the review guidelines that will help you to write reviews efficiently and effectively.
\nConfidentiality. By reviewing CleaR 2022, you must agree to keep all material and information related to the review confidential. In particular, you must not use ideas and results from submitted papers in your own research or distribute them to others. You should delete all reviewing material, such as the submitted code, at the end of the reviewing cycle. You should not talk about submissions or content related to the reviewing of submissions to anyone without prior approval from the program chairs.
\nDouble-blind reviewing. The CLeaR review process is double-blind: reviewers and authors will both stay anonymous to each other during the review process. However, author names will be visible to area chairs and program chairs. Authors are responsible for anonymizing their submissions. Submissions may not contain any identifying information that may violate the double-blind reviewing policy. If you are assigned a submission that is not adequately anonymized, please contact the corresponding AC. Also, you should not attempt to find out the identities of authors for any of your assigned submissions, e.g., by searching arXiv preprints. Reviewer names are visible to the area chair (and program chairs), but the reviewers will not know names of other reviewers. Please do not disclose your identity to authors and fellow reviewers in the discussions.
\nDual Submissions.CLeaR does not allow double submissions. Namely, submissions should not have been previously published in or submitted to a journal or the proceedings of another conference at any point during the CLeaR review process. Submissions as extended abstracts (5 pages or less), to workshops or non-archival venues (without a proceedings), will not be considered a concurrent submission. Authors may submit anonymized work to CLeaR that is already available as a preprint (e.g., on arXiv) without citing it. If you suspect that a submission that has been assigned to you is a dual submission or if you require further clarification, please contact the corresponding AC. Please see Call for Papers for more information about dual submissions.
\nViolations of formatting instructions. Submissions are limited to 12 single-column PMLR-formatted pages, plus unlimited additional pages for references and appendices. Authors of accepted papers will have the option of opting out of the proceedings in favor of a 1-page extended abstract, which will point to an open access archival version of the full paper reviewed for CLeaR. If you are assigned a paper that is overlength or appears to violate the CLeaR proceedings format (e.g., by decreasing margins or font size, by removing some pre-fixed spaces, etc), please notify the corresponding AC immediately.
\n \n* Please also review the policies in the CLeaR 2022 Call for Papers.
\n.default}) | \n .default}){kind=logo} | \n .default}) | \n
.default}) | \n .default}) | \n
.default}) | \n
\n Causality is a fundamental notion in science and engineering. In the\n past few decades, some of the most influential developments in the\n study of causal discovery, causal inference, and the causal treatment\n of machine learning have resulted from cross-disciplinary efforts. In\n particular, a number of machine learning and statistical analysis\n techniques have been developed to tackle classical causal discovery\n and inference problems. On the other hand, the causal view has been\n shown to be able to facilitate formulating, understanding, and\n tackling a number of hard machine learning problems in transfer\n learning, reinforcement learning, and deep learning.\n
\n\n We invite submissions to the 1st conference on Causal Learning and\n Reasoning (CLeaR), and welcome paper submissions that describe new\n theory, methodology, and/or applications relevant to any aspect of\n causal learning and reasoning in the fields of artificial intelligence\n and statistics. Submitted papers will be evaluated based on their\n novelty, technical quality, and potential impact. Experimental methods\n and results are expected to be reproducible, and authors are strongly\n encouraged to make code and data available. We also encourage\n submissions of proof-of-concept research that puts forward novel ideas\n and demonstrates potential for addressing problems at the intersection\n of causality and machine learning. CLeaR 2022 will be held in Eureka, CA, USA from April 11 to 13, 2022, with virtual elements. \n
\n Topics of submission may include, but are not limited to:\nSubmit at https://openreview.net/group?id=cclear.cc/CLeaR/2022/Conference.
\nCLeaR 2022 is co-located with a CIFAR/Sloan invitation-based causality workshop. This workshop is organized by CIFAR in conjunction with the Learning in Machines and Brains program directed by Yoshua Bengio and Yann LeCun.
\nWe invite submissions to the 1st Conference on Causal Learning and Reasoning (CLeaR), and welcome paper submissions that describe new theory, methodology, and/or applications relevant to any aspect of causal learning and reasoning in the fields of artificial intelligence and statistics. Accepted papers will be published in the Proceedings of Machine Learning Research (PMLR). \n
\n\tSubmit at https://openreview.net/group?id=cclear.cc/CLeaR/2022/Conference.
\n\tCausality is a fundamental notion in science and engineering. In the past few decades, some of the most influential developments in the study of causal discovery, causal inference, and the causal treatment of machine learning have resulted from cross-disciplinary efforts. In particular, a number of machine learning and statistical analysis techniques have been developed to tackle classical causal discovery and inference problems. On the other hand, the causal view has been shown to facilitate formulating, understanding, and tackling a broad range of problems, including domain generalization, robustness, trustworthiness, and fairness across machine learning, reinforcement learning, and statistics.
\n\n\tWe invite papers that describe new theory, methodology and/or applications relevant to any aspect of causal learning and reasoning in the fields of artificial intelligence and statistics. Submitted papers will be evaluated based on their novelty, technical quality, and potential impact. Experimental methods and results are expected to be reproducible, and authors are strongly encouraged to make code and data available. We also encourage submissions of proof-of-concept research that puts forward novel ideas and demonstrates potential for addressing problems at the intersection of causality and machine learning.
\n\t\n\tThe proceedings track is the standard CLeaR paper submission track. Papers will be selected via a rigorous double-blind peer-review process. All accepted papers will be presented at the Conference as contributed talks or as posters and will be published in the Proceedings.
\n\t\n\tTopics of submission may include, but are not limited to:\n\t
The CLeaR 2022 organizing committee prioritizes the safety and health of our community. We are still considering the format of the CLeaR 2022 conference. It will be preferably held as a hybrid conference with no mandatory physical attendance, but we also keep a backup plan of making the conference virtual in case of new pandemic situations. After our final decision, we will announce the format of the conference on the website. Thank you for your patience and understanding.
\n\n\tSubmissions are limited to 12 single-column PMLR-formatted pages, plus unlimited additional pages for references and appendices. Authors of accepted papers will have the option of opting out of the proceedings in favor of a 1-page extended abstract, which will point to an open access archival version of the full paper reviewed for CLeaR. You can also submit a single file of additional supplementary material separately, which may be either a pdf file (containing proof details, for instance) or a zip file that can include multiple files of all formats (such as code or videos). Note that reviewers are under no obligation to examine the supplementary material.
\n\n\tPlease format the paper using the official LaTeX style files. We do not support submission in formats other than LaTeX. Please do not modify the layout given by the style file.
\n\t\n\tSubmissions will be through OpenReview (https://openreview.net/group?id=cclear.cc/CLeaR/2022/Conference) and will be open approximately 4-6 weeks before the abstract submission deadline.
\n\n\tThe CLeaR review process is double-blind: reviewers and authors will both stay anonymous to each other during the review process. We use OpenReview to host papers; however, public discussions are not allowed during the review process. The review comments are only visible to program chairs, area chairs, and reviewers with submitted review comments. Papers will be desk-rejected if they contain any information that can violate the double-blind reviewing policy, such as the author names or their affiliations, acknowledgements, or links that can infer any author’s identity or institution. Self-citations are allowed as long as anonymity is preserved. It is up to the author’s discretion how best to preserve anonymity when including self-citations. Possibilities include: leaving out a self-citation, including it but replacing the citation text with “removed for anonymous submission,” or leaving the citation as-is. We recommend leaving in a moderate number of self-citations for published or otherwise well-known work.
\n\n\tRevisions are allowed in the submission system until the paper submission deadline. Changes will not be allowed afterwards.
\n\n\tWe strongly discourage advertising the preprint on social media or in the press while under submission to CLeaR. Preprints must not be explicitly identified as an CLeaR submission at any time during the review period (i.e., from the abstract submission deadline until the notification of the accept/reject decision).
\n\n\tCLeaR does not allow double submissions. Namely, submissions should not have been previously published in or submitted to a journal or the proceedings of another conference at any point during the CLeaR review process. Submissions to workshops or other non-archival venues (without a proceedings) will not be considered as dual submissions. Submissions as extended abstracts with 5 pages or less will not be considered a concurrent submission either. Authors may submit anonymized work to CLeaR that is already available as a preprint (e.g., on arXiv) without citing it. Tech reports (including reports on sites such as arXiv) do not count as prior publication. It is acceptable to have a substantially extended version of the submitted paper under consideration simultaneously for journal publication, so long as the journal version’s planned publication date is after our publication (April 13, 2022, tentatively), and it does not violate the journal's policy, the journal submission does not interfere with CLeaR right to publish the paper, and the situation is clearly described at the time of CLeaR submission. Please describe the situation in the appropriate box on the submission page (and do not include author information in the submission itself, to avoid accidental unblinding). Authors are also allowed to give talks to restricted audiences on the work(s) submitted to CLeaR during the review.
\n\n\tAll accepted papers will be presented at the Conference either as contributed talks or as posters, and will be published in the CLeaR Conference Proceedings in the Journal of Machine Learning Research Workshop and Conference Proceedings series. Papers for talks and posters will be treated equally in publication.
\n\n\tThe reviewers and area-chairs will have access to papers and supplementary materials that are assigned to them.
\n\t\n\tThe program chairs and workflow chairs will have access to all the papers. Everyone having access to papers and supplementary materials will be instructed to keep them confidential during the review process and delete them after the final decisions.
\n\t\n\tReviews will be visible to area chairs, program chairs, and workflow chairs throughout the process. At any stage of the process, author names will not be known to the reviewers or area chairs, but only visible to program chairs. Reviewer names are visible to the area chair (and program chairs), but the reviewers will not know names of other reviewers.
\n\n\tBernhard Schölkopf, Caroline Uhler & Kun ZhangThank you for serving as an area chair for CLeaR 2022. As an area chair, your job is to (1) ensure that all the submissions you are assigned have high-quality reviews and good discussions, and (2) write quality meta-reviews and make acceptance decisions. If there is any question, please contact the program chairs chairs.clear2022@gmail.com. The main tasks of area chairs are listed below.
\n* Please review the policies in the CLeaR 2022 Call for Papers and Reviewer Instructions.
\nAt CLeaR 2022, our goal is to create a community and environment that recognizes and respects the intrinsic value of everyone, which is essential for the open exchange of ideas, freedom of thought and expression, and respectful scientific debate at the conference.
\nAll participants, including but not limited to organizers, reviewers, speakers, sponsors, and volunteers at our Conference and Conference-sponsored social events are required to agree with this Code of Conduct both during an event and on official communication channels, including social media. In particular, sponsors should not use sexual, racial, or other offensive images, events, or other materials. This code applies to both official sponsors and any organization that uses the conference name as a brand as part of its activities during or around the conference.
\nCLeaR is committed to providing all participants with an experience free from harassment, bullying, discrimination and retaliation. This includes offensive comments related to age, gender, gender identity and expression, race, sexual orientation, physical or intellectual disability, physical appearance, body type, ethnicity, religion, politics, technical choices, or any other personal characteristics. No tolerance for bullying, intimidation, personal assault, harassment, continuous interruption of conversations or other activities, and behavior that interferes with the full participation of other participants. This includes sexual harassment, stalking, stalking, harassing photography or recording, inappropriate physical contact, unwelcome sexual attention, vulgar communication, and diminutive characteristics. The policies apply to actual meeting sites and conference venues, including both physical venues, online venues, and official virtual platforms, including but not limited to Openreview comments, video, virtual streaming, Q&A tools. For example, offensive or even threatening comments in Open review are prohibited. Also, Zoom bombing or any virtual activities that have nothing to do with the topic of discussion and that are detrimental to the purpose of the topic or program are not allowed.
\nParticipants who are asked by any member of the community to stop any improper behavior defined here should immediately comply. Meeting organizers may take further actions at their discretion, including: formally or informally warning offenders, expelling the meeting without refunds, refusing to submit or participate in future CLeaR meetings, reporting the incident to the offender’s local agency or funding agency, or reporting to the local Authorities or law enforcement agencies report the incident. The ''just kidding'' response is unacceptable. If action is taken, an appeal procedure will be provided.
\nIf you have any concerns about possible violation of the policies, please contact the conference chairs (chairs.clear2022@gmail.com) as soon as possible. Reports made during the conference will be responded to within 24 hours; those at other times in less than two weeks. Complaints and violations will be handled with discretion. We are ready and eager to help attendees contact relevant help services, escort them to a safe place, or to otherwise help people who have been harassed feel safe during the meeting. We gratefully accept feedback from the CLeaR community on our policy and actions.
\nA similar version has been used by other conferences, such as ICLR 2020 and UAI 2020.
\nThis page defines what constitutes a conflict of interest for the CLeaR 2022 review process.
\nIf you are an author, reviewer, or area chair, please make sure to create or update your OpenReview profile. You will be asked to declare two types of conflicts---domain conflicts and personal conflicts. Both types are declared by filling out appropriate sections of your OpenReview profile, as described below.
\nThe deadline for submitting your camera-ready version is Feb. 20, 2022. The camera-ready version of your paper may have one extra page, in order to address the review comments. That is, it may have 13 single-column PMLR-formatted pages, plus unlimited additional pages for references and appendices. Appendices should be included in the same pdf after the references section. Please use \\documentclass[final,12pt] {'{'}clear2022{'}'} when preparing the camera-ready version. Please don’t include any packages like “geometry” or “full page” to change the margins. For consistency across the proceedings, ensure your paper is formatted for letter paper rather than A4 (you can check with pdfinfo).
\n \nCLeaR 2022 will be held at the Sequoia Conference Center (SCC), Eureka, CA, USA on April 11-13, 2022. We encourage in-person participation, but the conference also has virtual elements. \n We are happy to announce that CLeaR will be co-located with a CIFAR/Sloan invitation-based causality workshop. This workshop is organized by CIFAR in conjunction with the Learning in Machines and Brains program directed by Yoshua Bengio and Yann LeCun. CLeaR participants are welcome to attend all sessions of the workshop as well.
\n\nPlease register for the CLeaR 2022 here (registration closed). Please let us know if you need letters for visa application.
\n\nWe would like to mention that a number of authors and a majority of the organizing committee (including Bernhard Schölkopf, Caroline Uhler, Kun Zhang, Francesco Locatello, Frederick Eberhardt, Daniel Malinsky, Sara Magliacane, Chandler Squires, Patrick Burauel, Elizabeth Ogburn, and Julius von Kügelgen) have expressed their intention to attend in person. Below is the nearby hotel information:
\n| Name | \nDistance to SCC | \nPrice (per room per night) | \nNumber of rooms (estimate) | \n
|---|---|---|---|
| Best Western Plus, Humboldt Bay Inn | \n1.5 miles | \n140 | \n10+ | \n
| Hotel Arcata | \n7.3 miles | \n220+ | \n20 | \n
| Holiday Inn Express & Suites | \n2.1 miles | \n150+ | \n10+ | \n
| Clarion Hotel by Humboldt Bay | \n0.4 miles | \n112+ | \n35+ | \n
| The Inn at 2nd & C | \n1.4 miles | \n149+ | \n15+ | \n
| Carter House Inns | \n0.8 miles | \n225 | \n15+ | \n
| Name | \nDate | \n
|---|---|
| {name} | \n{date} | \n
\n {position}\n {persons.map(({ name, href, address }) => (\n \n ))}\n | \n
Thank you for agreeing to review CLearR 2023! Your assessment is vital to creating a high quality program. This page provides the review guidelines that will help you to write reviews efficiently and effectively.
\nConfidentiality. By reviewing CleaR 2023, you must agree to keep all material and information related to the review confidential. In particular, you must not use ideas and results from submitted papers in your own research or distribute them to others. You should delete all reviewing material, such as the submitted code, at the end of the reviewing cycle. You should not talk about submissions or content related to the reviewing of submissions to anyone without prior approval from the program chairs.
\nDouble-blind reviewing. The CLeaR review process is double-blind: reviewers and authors will both stay anonymous to each other during the review process. However, author names will be visible to area chairs and program chairs. Authors are responsible for anonymizing their submissions. Submissions may not contain any identifying information that may violate the double-blind reviewing policy. If you are assigned a submission that is not adequately anonymized, please contact the corresponding AC. Also, you should not attempt to find out the identities of authors for any of your assigned submissions, e.g., by searching arXiv preprints. Reviewer names are visible to the area chair (and program chairs), but the reviewers will not know names of other reviewers. Please do not disclose your identity to authors and fellow reviewers in the discussions.
\nDual Submissions.CLeaR does not allow double submissions. Namely, submissions should not have been previously published in or submitted to a journal or the proceedings of another conference at any point during the CLeaR review process. Submissions as extended abstracts (5 pages or less), to workshops or non-archival venues (without a proceedings), will not be considered a concurrent submission. Authors may submit anonymized work to CLeaR that is already available as a preprint (e.g., on arXiv) without citing it. If you suspect that a submission that has been assigned to you is a dual submission or if you require further clarification, please contact the corresponding AC. Please see Call for Papers for more information about dual submissions.
\nViolations of formatting instructions. Submissions are limited to 12 single-column PMLR-formatted pages, plus unlimited additional pages for references and appendices. Authors of accepted papers will have the option of opting out of the proceedings in favor of a 1-page extended abstract, which will point to an open access archival version of the full paper reviewed for CLeaR. If you are assigned a paper that is overlength or appears to violate the CLeaR proceedings format (e.g., by decreasing margins or font size, by removing some pre-fixed spaces, etc), please notify the corresponding AC immediately.
\n \n* Please also review the policies in the CLeaR 2023 Call for Papers.
\n | \n {/* | */}\n .default}) | \n
| */}\n .default}) | \n
| \n
\n Causality is a fundamental notion in science and engineering. In the\n past few decades, some of the most influential developments in the\n study of causal discovery, causal inference, and the causal treatment\n of machine learning have resulted from cross-disciplinary efforts. In\n particular, a number of machine learning and statistical analysis\n techniques have been developed to tackle classical causal discovery\n and inference problems. On the other hand, the causal view has been\n shown to be able to facilitate formulating, understanding, and\n tackling a number of hard machine learning problems in transfer\n learning, reinforcement learning, and deep learning.\n
\n\n CLeaR 2022: Starting a brand new conference in these pandemic years\n and ensuring it is set up for long-term success has been a significant\n undertaking. Despite these challenges, more than 50 people attended the\n conference in person and several hundred connected remotely. We had 9\n oral presentations and 40 posters, covering topics that range from causal\n discovery, causal fairness, explainability, non-parametric inference,\n causal Markov decision processes, to social-influence estimation,\n applications of causality, and other topics. We have received a number\n of enquiries about whether and where to hold CLeaR 2023 and are\n delighted to announce the next edition.\n
\n\n We invite submissions to the 2nd conference on Causal Learning and\n Reasoning (CLeaR), and welcome paper submissions that describe new\n theory, methodology, and/or applications relevant to any aspect of\n causal learning and reasoning in the fields of artificial intelligence\n and statistics. Submitted papers will be evaluated based on their\n novelty, technical quality, and potential impact. Experimental methods\n and results are expected to be reproducible, and authors are strongly\n encouraged to make code and data available. We also encourage\n submissions of proof-of-concept research that puts forward novel ideas\n and demonstrates potential for addressing problems at the intersection\n of causality and machine learning. CLeaR 2023 will be held in Tübingen,\n Germany from April 11 to 14, 2023, with virtual elements. \n
\n Topics of submission may include, but are not limited to:\nSubmit at https://openreview.net/group?id=cclear.cc/CLeaR/2023/Conference.
\nThe table discussion results on Causality in the time of LLMs are released. PLease find the details here
\nWe invite submissions to the 2nd Conference on Causal Learning and Reasoning (CLeaR), and welcome paper submissions that describe new theory, methodology, and/or applications relevant to any aspect of causal learning and reasoning in the fields of artificial intelligence and statistics. Accepted papers will be published in the Proceedings of Machine Learning Research (PMLR). \n
\n\tSubmit at https://openreview.net/group?id=cclear.cc/CLeaR/2023/Conference.
\n\tCausality is a fundamental notion in science and engineering. In the past few decades, some of the most influential developments in the study of causal discovery, causal inference, and the causal treatment of machine learning have resulted from cross-disciplinary efforts. In particular, a number of machine learning and statistical analysis techniques have been developed to tackle classical causal discovery and inference problems. On the other hand, the causal view has been shown to facilitate formulating, understanding, and tackling a broad range of problems, including domain generalization, robustness, trustworthiness, and fairness across machine learning, reinforcement learning, and statistics.
\n\n\tWe invite papers that describe new theory, methodology and/or applications relevant to any aspect of causal learning and reasoning in the fields of artificial intelligence and statistics. Submitted papers will be evaluated based on their novelty, technical quality, and potential impact. Experimental methods and results are expected to be reproducible, and authors are strongly encouraged to make code and data available. We also encourage submissions of proof-of-concept research that puts forward novel ideas and demonstrates potential for addressing problems at the intersection of causality and machine learning.
\n\t\n\tThe proceedings track is the standard CLeaR paper submission track. Papers will be selected via a rigorous double-blind peer-review process. All accepted papers will be presented at the Conference as contributed talks or as posters and will be published in the Proceedings.
\n\t\n\tTopics of submission may include, but are not limited to:\n\t
The CLeaR 2023 organizing committee prioritizes the safety and health of our community. We are still considering the format of the CLeaR 2023 conference. It will be preferably held as a hybrid conference with no mandatory physical attendance, but we also keep a backup plan of making the conference virtual in case of new pandemic situations. After our final decision, we will announce the format of the conference on the website. Thank you for your patience and understanding.
\n\n\tSubmissions are limited to 12 single-column PMLR-formatted pages, plus unlimited additional pages for references and appendices. Authors of accepted papers will have the option of opting out of the proceedings in favor of a 1-page extended abstract, which will point to an open access archival version of the full paper reviewed for CLeaR. You can also submit a single file of additional supplementary material separately, which may be either a pdf file (containing proof details, for instance) or a zip file that can include multiple files of all formats (such as code or videos). Note that reviewers are under no obligation to examine the supplementary material.
\n\n\tPlease format the paper using the official LaTeX style files. We do not support submission in formats other than LaTeX. Please do not modify the layout given by the style file.
\n\t\n\tSubmissions will be through OpenReview (https://openreview.net/group?id=cclear.cc/CLeaR/2023/Conference) and will be open approximately 4-6 weeks before the paper submission deadline.
\n\n\tThe CLeaR review process is double-blind: reviewers and authors will both stay anonymous to each other during the review process. We use OpenReview to host papers; however, public discussions are not allowed during the review process. The review comments are only visible to program chairs, area chairs, and reviewers with submitted review comments. Papers will be desk-rejected if they contain any information that can violate the double-blind reviewing policy, such as the author names or their affiliations, acknowledgements, or links that can infer any author’s identity or institution. Self-citations are allowed as long as anonymity is preserved. It is up to the author’s discretion how best to preserve anonymity when including self-citations. Possibilities include: leaving out a self-citation, including it but replacing the citation text with “removed for anonymous submission,” or leaving the citation as-is. We recommend leaving in a moderate number of self-citations for published or otherwise well-known work.
\n\n\tRevisions are allowed in the submission system until the paper submission deadline. Changes will not be allowed afterwards.
\n\n\tWe strongly discourage advertising the preprint on social media or in the press while under submission to CLeaR. Preprints must not be explicitly identified as an CLeaR submission at any time during the review period (i.e., from the abstract submission deadline until the notification of the accept/reject decision).
\n\n\tCLeaR does not allow double submissions. Namely, submissions should not have been previously published in or submitted to a journal or the proceedings of another conference at any point during the CLeaR review process. Submissions to workshops or other non-archival venues (without a proceedings) will not be considered as dual submissions. Submissions as extended abstracts with 5 pages or less will not be considered a concurrent submission either. Authors may submit anonymized work to CLeaR that is already available as a preprint (e.g., on arXiv) without citing it. Tech reports (including reports on sites such as arXiv) do not count as prior publication. It is acceptable to have a substantially extended version of the submitted paper under consideration simultaneously for journal publication, so long as the journal version’s planned publication date is after our publication (April 13, 2023, tentatively), and it does not violate the journal's policy, the journal submission does not interfere with CLeaR right to publish the paper, and the situation is clearly described at the time of CLeaR submission. Please describe the situation in the appropriate box on the submission page (and do not include author information in the submission itself, to avoid accidental unblinding). Authors are also allowed to give talks to restricted audiences on the work(s) submitted to CLeaR during the review.
\n\n\tAll accepted papers will be presented at the Conference either as contributed talks or as posters, and will be published in the CLeaR Conference Proceedings in the Journal of Machine Learning Research Workshop and Conference Proceedings series. Papers for talks and posters will be treated equally in publication.
\n\n\tThe reviewers and area-chairs will have access to papers and supplementary materials that are assigned to them.
\n\t\n\tThe program chairs and workflow chairs will have access to all the papers. Everyone having access to papers and supplementary materials will be instructed to keep them confidential during the review process and delete them after the final decisions.
\n\t\n\tReviews will be visible to area chairs, program chairs, and workflow chairs throughout the process. At any stage of the process, author names will not be known to the reviewers or area chairs, but only visible to program chairs. Reviewer names are visible to the area chair (and program chairs), but the reviewers will not know names of other reviewers.
\n\n\tMihaela van der Shaar, Cheng Zhang & Dominik Janzing
\n\tCLeaR 2023 Program Chairs
Francesco Locatello & Peter Spirtes
\n\tCLeaR 2023 General Chairs
We are thrilled to announce our causal benchmark dataset track as part of our CLeaR 2023 conference. High-quality, publicly available causal datasets are critical for advancing the field of causality, and we seek to provide researchers with a diverse collection of datasets that can be routinely used to test, benchmark, and improve the overall performance of causal methods and algorithms. This includes but is not limited to causal inference, causal discovery, and causal representation learning. By submitting your dataset to the CLeaR 2023 dataset track, you will not only be contributing to enabling new discoveries in the field of causality, but also empowering the real-world impact in various domains such as natural sciences, healthcare, social sciences, and economics. Together with the various research fields, we hope to have a positive impact on the future of the field, and to advance our understand of causality. We encourage researchers from all fields to submit their datasets and be part of this exciting initiative.
\n\tWe hope to make this initiative as inclusive as possible, therefore datasets with diversity in terms of domain, complexity, and scale are all welcome to submit. Submissions should include a clear description of the underlying problem setting, data format, causal quantities of interest, as well as clear definition of evaluation metrics. Submissions are NOT required to be anonymized. All submissions should comply with the TMLR guidelines for “General Ethical Conduct” outlined here. We note that all submissions will receive a light touch review by committee members to verify that they comply with the guidelines above and ensure the data provided will be of value to the causality community.
\n\nNon-archival policy. The dataset track at CLEAR is a non-archival venue and will not have official proceedings. Submissions to the dataset track can be subsequently or concurrently submitted to other venues, but will be hosted on a dedicated website (and indexed by Google Scholar).
\n\nTo submit a dataset, please send the following material to our official email (clear.datasets.2023@gmail.com):
\n\tFormal review process will not be conducted for submissions. Instead, this initiative will contain the following phases:
\nThank you for serving as an area chair for CLeaR 2023. As an area chair, your job is to (1) ensure that all the submissions you are assigned have high-quality reviews and good discussions, and (2) write quality meta-reviews and make acceptance decisions. If there is any question, please contact the program chairs chairs.clear2023@gmail.com. The main tasks of area chairs are listed below.
\n* Please review the policies in the CLeaR 2023 Call for Papers and Reviewer Instructions.
\nAt CLeaR 2023, our goal is to create a community and environment that recognizes and respects the intrinsic value of everyone, which is essential for the open exchange of ideas, freedom of thought and expression, and respectful scientific debate at the conference.
\nAll participants, including but not limited to organizers, reviewers, speakers, sponsors, and volunteers at our Conference and Conference-sponsored social events are required to agree with this Code of Conduct both during an event and on official communication channels, including social media. In particular, sponsors should not use sexual, racial, or other offensive images, events, or other materials. This code applies to both official sponsors and any organization that uses the conference name as a brand as part of its activities during or around the conference.
\nCLeaR is committed to providing all participants with an experience free from harassment, bullying, discrimination and retaliation. This includes offensive comments related to age, gender, gender identity and expression, race, sexual orientation, physical or intellectual disability, physical appearance, body type, ethnicity, religion, politics, technical choices, or any other personal characteristics. No tolerance for bullying, intimidation, personal assault, harassment, continuous interruption of conversations or other activities, and behavior that interferes with the full participation of other participants. This includes sexual harassment, stalking, stalking, harassing photography or recording, inappropriate physical contact, unwelcome sexual attention, vulgar communication, and diminutive characteristics. The policies apply to actual meeting sites and conference venues, including both physical venues, online venues, and official virtual platforms, including but not limited to Openreview comments, video, virtual streaming, Q&A tools. For example, offensive or even threatening comments in Open review are prohibited. Also, Zoom bombing or any virtual activities that have nothing to do with the topic of discussion and that are detrimental to the purpose of the topic or program are not allowed.
\nParticipants who are asked by any member of the community to stop any improper behavior defined here should immediately comply. Meeting organizers may take further actions at their discretion, including: formally or informally warning offenders, expelling the meeting without refunds, refusing to submit or participate in future CLeaR meetings, reporting the incident to the offender’s local agency or funding agency, or reporting to the local Authorities or law enforcement agencies report the incident. The ''just kidding'' response is unacceptable. If action is taken, an appeal procedure will be provided.
\nIf you have any concerns about possible violation of the policies, please contact the conference chairs (chairs.clear2023@gmail.com) as soon as possible. Reports made during the conference will be responded to within 24 hours; those at other times in less than two weeks. Complaints and violations will be handled with discretion. We are ready and eager to help attendees contact relevant help services, escort them to a safe place, or to otherwise help people who have been harassed feel safe during the meeting. We gratefully accept feedback from the CLeaR community on our policy and actions.
\nA similar version has been used by other conferences, such as ICLR 2020 and UAI 2020.
\nThis page defines what constitutes a conflict of interest for the CLeaR 2023 review process.
\nIf you are an author, reviewer, or area chair, please make sure to create or update your OpenReview profile. You will be asked to declare two types of conflicts---domain conflicts and personal conflicts. Both types are declared by filling out appropriate sections of your OpenReview profile, as described below.
\nTBD
\n \nCLeaR 2023 will be held at the Amazon Development Center, Tübingen, Germany on April 11-14, 2023. We encourage in-person participation, but the conference also has virtual elements.
\nPlease register for the CLeaR 2023 here (registration closes March 15). Please let us know if you need letters for visa application.
\nPlease check out the description of the Max-Planck-Institute for Intelligent Systems on how to get to Tuebingen using public transport here.
\nYou will receive a public transport ticket during your registration, which is valid during the conference period (11th to the 14th), but only for the city busses in Tuebingen (it is not valid all the way to Stuttgart or the airport). Note: You need to get to the registration by buying a one-way ticket!
\nGoogle maps’ public transport suggestions work fairly well in Tuebingen. There are multiple bus lines that bring you to the venue or close to it.
\nThe easiest is bus line 3 towards Waldhäuser Ost, get off the bus at “Max-Planck-Institute” and you will see the Amazon building right away.
\n.default})
\n Other options are bus lines 4 or 6, also towards Waldhäuser Ost, bring you close to the venue as well. However, you will need to walk from there for up to 10 minutes to get to the venue (get off at “Winkelwiese”).
\n.default})
\n Bus line 17 also works (get off at “Robert-Gradmann-Weg” and continue walking up the hill).
\n.default})
\n We do not partner with any hotel in Tuebingen. Here is an incomplete list:
\n \nWe highly recommend to attend the conference in-person, even more so for presenting your accepted papers.
\nPlease contact us at zietld@amazon.com in case you cannot present (oral or poster) in-person (subject: [CleaR 2023 remote]).
\nTime: {time}
\nJoin Zoom Meeting:
\n \nMeeting ID: {meetingId}
\nPasscode: {passcode}
\n \nRegistered participants can attend keynotes and oral presentations (see schedule) remotely. Please connect to this Zoom meeting:
\nThis link is used for all keynotes and orals. Note that in-person poster sessions, the townhall, panel discussion, and social activities will not be streamed.
\nWhen participating remotely, please mute yourself (except when asking questions). Feel free to ask questions either in the chat or by raising your hand in Zoom and asking them over video.
\nVirtual posters can be found below.
\n| Paper title | \nDownload link | \n
|---|---|
| {paper.title} | \nDownload | \n
| What | \nWhen | \n
|---|---|
| {name} | \n{date} | \n
\n {position}\n {persons.map(({ name, href, address }) => (\n \n ))}\n | \n
Thank you for agreeing to review CLearR 2024! Your assessment is vital to creating a high quality program. This page provides the review guidelines that will help you to write reviews efficiently and effectively.
\nConfidentiality. By reviewing CleaR 2024, you must agree to keep all material and information related to the review confidential. In particular, you must not use ideas and results from submitted papers in your own research or distribute them to others. You should delete all reviewing material, such as the submitted code, at the end of the reviewing cycle. You should not talk about submissions or content related to the reviewing of submissions to anyone without prior approval from the program chairs.
\nDouble-blind reviewing. The CLeaR review process is double-blind: reviewers and authors will both stay anonymous to each other during the review process. However, author names will be visible to area chairs and program chairs. Authors are responsible for anonymizing their submissions. Submissions may not contain any identifying information that may violate the double-blind reviewing policy. If you are assigned a submission that is not adequately anonymized, please contact the corresponding AC. Also, you should not attempt to find out the identities of authors for any of your assigned submissions, e.g., by searching arXiv preprints. Reviewer names are visible to the area chair (and program chairs), but the reviewers will not know names of other reviewers. Please do not disclose your identity to authors and fellow reviewers in the discussions.
\nDual Submissions.CLeaR does not allow double submissions. Namely, submissions should not have been previously published in or submitted to a journal or the proceedings of another conference at any point during the CLeaR review process. Submissions as extended abstracts (5 pages or less), to workshops or non-archival venues (without a proceedings), will not be considered a concurrent submission. Authors may submit anonymized work to CLeaR that is already available as a preprint (e.g., on arXiv) without citing it. If you suspect that a submission that has been assigned to you is a dual submission or if you require further clarification, please contact the corresponding AC. Please see Call for Papers for more information about dual submissions.
\nViolations of formatting instructions. Submissions are limited to 12 single-column PMLR-formatted pages, plus unlimited additional pages for references and appendices. Authors of accepted papers will have the option of opting out of the proceedings in favor of a 1-page extended abstract, which will point to an open access archival version of the full paper reviewed for CLeaR. If you are assigned a paper that is overlength or appears to violate the CLeaR proceedings format (e.g., by decreasing margins or font size, by removing some pre-fixed spaces, etc), please notify the corresponding AC immediately.
\n \n* Please also review the policies in the CLeaR 2024 Call for Papers.
\n.default}) | \n
| */}\n .default}){kind=logo} | \n
.default}){kind=logo} | \n
TBD
\n\n Causality is a fundamental notion in science and engineering. In the past few decades, some of the most influential developments in the study of causal discovery, causal inference, and the causal treatment of machine learning have resulted from cross-disciplinary efforts. In particular, a number of machine learning and statistical analysis techniques have been developed to tackle classical causal discovery and inference problems. On the other hand, the causal view has been shown to be able to facilitate formulating, understanding, and tackling a number of hard machine learning problems in transfer learning, reinforcement learning, and deep learning.\n
\n\n We invite submissions to the 3rd conference on Causal Learning and Reasoning (CLeaR), and welcome paper submissions that describe new theory, methodology, and/or applications relevant to any aspect of causal learning and reasoning in the fields of artificial intelligence and statistics. Submitted papers will be evaluated based on their novelty, technical quality, and potential impact. Experimental methods and results are expected to be reproducible, and authors are strongly encouraged to make code and data available. We also encourage submissions of proof-of-concept research that puts forward novel ideas and demonstrates potential for addressing problems at the intersection of causality and machine learning. \n CLeaR 2024 will be held at the Pallisades Room, 3rd Floor, Carnesale Commons, UC Los Angeles, California from April 1 to 3, 2024.\n
\n Topics of submission may include, but are not limited to:\nDifferent from last year, there will be a single track for all submissions (no separate call for dataset submissions). The program at the conference will highlight new applications of causality to real-world problems, benchmark datasets, or benchmark methodologies.
\nWe invite submissions to the 3rd Conference on Causal Learning and Reasoning (CLeaR), and welcome paper submissions that describe new theory, methodology, and/or applications relevant to any aspect of causal learning and reasoning in the fields of artificial intelligence and statistics. Accepted papers will be published in the Proceedings of Machine Learning Research (PMLR). \n
\n\tSubmit at https://openreview.net/group?id=cclear.cc/CLeaR/2024/Conference.
\n\n\tCausality is a fundamental notion in science and engineering. In the past few decades, some of the most influential developments in the study of causal discovery, causal inference, and the causal treatment of machine learning have resulted from cross-disciplinary efforts. In particular, a number of machine learning and statistical analysis techniques have been developed to tackle classical causal discovery and inference problems. On the other hand, the causal view has been shown to facilitate formulating, understanding, and tackling a broad range of problems, including domain generalization, robustness, trustworthiness, and fairness across machine learning, reinforcement learning, and statistics.
\n\n\tWe invite papers that describe new theory, methodology and/or applications relevant to any aspect of causal learning and reasoning in the fields of artificial intelligence and statistics. Submitted papers will be evaluated based on their novelty, technical quality, and potential impact. Experimental methods and results are expected to be reproducible, and authors are strongly encouraged to make code and data available. We also encourage submissions of proof-of-concept research that puts forward novel ideas and demonstrates potential for addressing problems at the intersection of causality and machine learning.
\n\t\n\tThe proceedings track is the standard CLeaR paper submission track. Papers will be selected via a rigorous double-blind peer-review process. All accepted papers will be presented at the Conference as contributed talks or as posters and will be published in the Proceedings.
\n\t\n\tTopics of submission may include, but are not limited to:\n\t
The CLeaR 2024 organizing committee prioritizes the safety and health of our community. We are still considering the format of the CLeaR 2024 conference. It will be preferably held as a hybrid conference with no mandatory physical attendance, but we also keep a backup plan of making the conference virtual in case of new pandemic situations. After our final decision, we will announce the format of the conference on the website. Thank you for your patience and understanding.
\n\n\tSubmissions are limited to 12 single-column PMLR-formatted pages, plus unlimited additional pages for references and appendices. Authors of accepted papers will have the option of opting out of the proceedings in favor of a 1-page extended abstract, which will point to an open access archival version of the full paper reviewed for CLeaR. You can also submit a single file of additional supplementary material separately, which may be either a pdf file (containing proof details, for instance) or a zip file that can include multiple files of all formats (such as code or videos). Note that reviewers are under no obligation to examine the supplementary material.
\n\n\tPlease format the paper using the official LaTeX style files. We do not support submission in formats other than LaTeX. Please do not modify the layout given by the style file.
\n\n\n\tThe CLeaR review process is double-blind: reviewers and authors will both stay anonymous to each other during the review process. We use OpenReview to host papers; however, public discussions are not allowed during the review process. The review comments are only visible to program chairs, area chairs, and reviewers with submitted review comments. Papers will be desk-rejected if they contain any information that can violate the double-blind reviewing policy, such as the author names or their affiliations, acknowledgements, or links that can infer any author’s identity or institution. Self-citations are allowed as long as anonymity is preserved. It is up to the author’s discretion how best to preserve anonymity when including self-citations. Possibilities include: leaving out a self-citation, including it but replacing the citation text with “removed for anonymous submission,” or leaving the citation as-is. We recommend leaving in a moderate number of self-citations for published or otherwise well-known work.
\n\n\tRevisions are allowed in the submission system until the paper submission deadline. Changes will not be allowed afterwards.
\n\n\tWe strongly discourage advertising the preprint on social media or in the press while under submission to CLeaR. Preprints must not be explicitly identified as an CLeaR submission at any time during the review period (i.e., from the abstract submission deadline until the notification of the accept/reject decision).
\n\n\tCLeaR does not allow double submissions. Namely, submissions should not have been previously published in or submitted to a journal or the proceedings of another conference at any point during the CLeaR review process. Submissions to workshops or other non-archival venues (without a proceedings) will not be considered as dual submissions. Submissions as extended abstracts with 5 pages or less will not be considered a concurrent submission either. Authors may submit anonymized work to CLeaR that is already available as a preprint (e.g., on arXiv) without citing it. Tech reports (including reports on sites such as arXiv) do not count as prior publication. It is acceptable to have a substantially extended version of the submitted paper under consideration simultaneously for journal publication, so long as the journal version’s planned publication date is after our publication (April 13, 2024, tentatively), and it does not violate the journal's policy, the journal submission does not interfere with CLeaR right to publish the paper, and the situation is clearly described at the time of CLeaR submission. Please describe the situation in the appropriate box on the submission page (and do not include author information in the submission itself, to avoid accidental unblinding). Authors are also allowed to give talks to restricted audiences on the work(s) submitted to CLeaR during the review.
\n\n\tAll accepted papers will be presented at the Conference either as contributed talks or as posters, and will be published in the CLeaR Conference Proceedings in the Journal of Machine Learning Research Workshop and Conference Proceedings series. Papers for talks and posters will be treated equally in publication.
\n\n\tThe reviewers and area-chairs will have access to papers and supplementary materials that are assigned to them.
\n\t\n\tThe program chairs and workflow chairs will have access to all the papers. Everyone having access to papers and supplementary materials will be instructed to keep them confidential during the review process and delete them after the final decisions.
\n\t\n\tReviews will be visible to area chairs, program chairs, and workflow chairs throughout the process. At any stage of the process, author names will not be known to the reviewers or area chairs, but only visible to program chairs. Reviewer names are visible to the area chair (and program chairs), but the reviewers will not know names of other reviewers.
\n\n\tVanessa Didelez & Francesco Locatello
\n\tCLeaR 2024 Program Chairs
Aditya Grover & Cheng Zhang
\n\tCLeaR 2024 General Chairs
Thank you for serving as an area chair for CLeaR 2024. As an area chair, your job is to (1) ensure that all the submissions you are assigned have high-quality reviews and good discussions, and (2) write quality meta-reviews and make acceptance decisions. If there is any question, please contact the program chairs chairs.clear2024@gmail.com. The main tasks of area chairs are listed below.
\n* Please review the policies in the CLeaR 2024 Call for Papers and Reviewer Instructions.
\nAt CLeaR 2023, our goal is to create a community and environment that recognizes and respects the intrinsic value of everyone, which is essential for the open exchange of ideas, freedom of thought and expression, and respectful scientific debate at the conference.
\nAll participants, including but not limited to organizers, reviewers, speakers, sponsors, and volunteers at our Conference and Conference-sponsored social events are required to agree with this Code of Conduct both during an event and on official communication channels, including social media. In particular, sponsors should not use sexual, racial, or other offensive images, events, or other materials. This code applies to both official sponsors and any organization that uses the conference name as a brand as part of its activities during or around the conference.
\nCLeaR is committed to providing all participants with an experience free from harassment, bullying, discrimination and retaliation. This includes offensive comments related to age, gender, gender identity and expression, race, sexual orientation, physical or intellectual disability, physical appearance, body type, ethnicity, religion, politics, technical choices, or any other personal characteristics. No tolerance for bullying, intimidation, personal assault, harassment, continuous interruption of conversations or other activities, and behavior that interferes with the full participation of other participants. This includes sexual harassment, stalking, stalking, harassing photography or recording, inappropriate physical contact, unwelcome sexual attention, vulgar communication, and diminutive characteristics. The policies apply to actual meeting sites and conference venues, including both physical venues, online venues, and official virtual platforms, including but not limited to Openreview comments, video, virtual streaming, Q&A tools. For example, offensive or even threatening comments in Open review are prohibited. Also, Zoom bombing or any virtual activities that have nothing to do with the topic of discussion and that are detrimental to the purpose of the topic or program are not allowed.
\nParticipants who are asked by any member of the community to stop any improper behavior defined here should immediately comply. Meeting organizers may take further actions at their discretion, including: formally or informally warning offenders, expelling the meeting without refunds, refusing to submit or participate in future CLeaR meetings, reporting the incident to the offender’s local agency or funding agency, or reporting to the local Authorities or law enforcement agencies report the incident. The ''just kidding'' response is unacceptable. If action is taken, an appeal procedure will be provided.
\nIf you have any concerns about possible violation of the policies, please contact the conference chairs (chairs.clear2023@gmail.com) as soon as possible. Reports made during the conference will be responded to within 24 hours; those at other times in less than two weeks. Complaints and violations will be handled with discretion. We are ready and eager to help attendees contact relevant help services, escort them to a safe place, or to otherwise help people who have been harassed feel safe during the meeting. We gratefully accept feedback from the CLeaR community on our policy and actions.
\nA similar version has been used by other conferences, such as ICLR 2020 and UAI 2020.
\nThis page defines what constitutes a conflict of interest for the CLeaR 2024 review process.
\nIf you are an author, reviewer, or area chair, please make sure to create or update your OpenReview profile. You will be asked to declare two types of conflicts---domain conflicts and personal conflicts. Both types are declared by filling out appropriate sections of your OpenReview profile, as described below.
\nTBD
\n \nCLeaR 2024 will be held at Carnesale Commons, UC Los Angeles, California from April 1 to 3, 2024. We encourage participation in person, but the conference also includes virtual components.
\nYou can register for CLeaR 2024 here. Registration will be open until March 15, 2024. As we have a limited number of available tickets, please register as soon as possible. Please contact us if you need an invitation letter for a visa application.
\nWe highly recommend to attend the conference in-person, even more so for presenting your accepted papers.
\nPlease contact us at zietld@amazon.com in case you cannot present (oral or poster) in-person (subject: [CleaR 2023 remote]).
\nTime: {time}
\nJoin Zoom Meeting:
\n \nMeeting ID: {meetingId}
\nPasscode: {passcode}
\n \nRegistered participants can attend keynotes and oral presentations (see schedule) remotely. Please connect to this Zoom meeting:
\nThis link is used for all keynotes and orals. Note that in-person poster sessions, the townhall, panel discussion, and social activities will not be streamed.
\nWhen participating remotely, please mute yourself (except when asking questions). Feel free to ask questions either in the chat or by raising your hand in Zoom and asking them over video.
\nVirtual posters can be found below.
\n| Paper title | \nDownload link | \n
|---|---|
| {paper.title} | \nDownload | \n
TBD
\n| What | \nWhen | \n
|---|---|
| {name} | \n{date} | \n
\n {position}\n {persons.map(({ name, href, address }) => (\n \n ))}\n | \n
Thank you for agreeing to review CLeaR 2025! Your assessment is vital to creating a high quality program. This page provides the review guidelines that will help you to write reviews efficiently and effectively.
\nConfidentiality. By reviewing CLeaR 2025, you must agree to keep all material and information related to the review confidential. In particular, you must not use ideas and results from submitted papers in your own research or distribute them to others. You should delete all reviewing material, such as the submitted code, at the end of the reviewing cycle. You should not talk about submissions or content related to the reviewing of submissions to anyone without prior approval from the program chairs.
\nDouble-blind reviewing. The CLeaR review process is double-blind: reviewers and authors will both stay anonymous to each other during the review process. However, author names will be visible to area chairs and program chairs. Authors are responsible for anonymizing their submissions. Submissions may not contain any identifying information that may violate the double-blind reviewing policy. If you are assigned a submission that is not adequately anonymized, please contact the corresponding AC. Also, you should not attempt to find out the identities of authors for any of your assigned submissions, e.g., by searching arXiv preprints. Reviewer names are visible to the area chair (and program chairs), but the reviewers will not know names of other reviewers. Please do not disclose your identity to authors and fellow reviewers in the discussions.
\nDual Submissions.CLeaR does not allow double submissions. Namely, submissions should not have been previously published in or submitted to a journal or the proceedings of another conference at any point during the CLeaR review process. Submissions as extended abstracts (5 pages or less), to workshops or non-archival venues (without a proceedings), will not be considered a concurrent submission. Authors may submit anonymized work to CLeaR that is already available as a preprint (e.g., on arXiv) without citing it. If you suspect that a submission that has been assigned to you is a dual submission or if you require further clarification, please contact the corresponding AC. Please see Call for Papers for more information about dual submissions.
\nViolations of formatting instructions. Submissions are limited to 12 single-column PMLR-formatted pages, plus unlimited additional pages for references and appendices. Authors of accepted papers will have the option of opting out of the proceedings in favor of a 1-page extended abstract, which will point to an open access archival version of the full paper reviewed for CLeaR. If you are assigned a paper that is overlength or appears to violate the CLeaR proceedings format (e.g., by decreasing margins or font size, by removing some pre-fixed spaces, etc), please notify the corresponding AC immediately.
\n\n* Please also review the policies in the CLeaR 2025 Call for Papers.
\nTBA
\n\n Causality is a fundamental notion in science and engineering. In the past few decades, some of the most influential developments in the study of causal discovery, causal inference, and the causal treatment of machine learning have resulted from cross-disciplinary efforts. In particular, a number of machine learning and statistical analysis techniques have been developed to tackle classical causal discovery and inference problems. On the other hand, the causal view has been shown to be able to facilitate formulating, understanding, and tackling a number of hard machine learning problems in transfer learning, reinforcement learning, and deep learning.\n
\n\n We invite submissions to the 4th conference on Causal Learning and Reasoning (CLeaR),\n and welcome paper submissions that describe new theory, methodology, and/or applications relevant to any aspect of causal\n learning and reasoning in the fields of artificial intelligence and statistics. Submitted papers will be evaluated based\n on their novelty, technical quality, and potential impact. Experimental methods and results are expected to be reproducible,\n and authors are strongly encouraged to make code and data available.\n We also encourage submissions of proof-of-concept research that puts forward novel ideas and demonstrates potential for\n addressing problems at the intersection of causality and machine learning.\n
\nCLeaR 2025 will be held in Lausanne, Switzerland, May 7 to May 9, 2025.
\n Topics of submission may include, but are not limited to:\nThe program at the conference will highlight new applications of causality to real-world problems, benchmark datasets,\n or benchmark methodologies.
\nBelow please find the schedule of the CLeaR 2025 conference:
\n\nWe invite submissions to the 4th Conference on Causal Learning and Reasoning (CLeaR), and welcome paper submissions that describe new theory, methodology, and/or applications relevant to any aspect of causal learning and reasoning in the fields of artificial intelligence and statistics. Accepted papers will be published in the Proceedings of Machine Learning Research (PMLR).\n
\nSubmit at https://openreview.net/group?id=cclear.cc/CLeaR/2025/Conference.
\n\nCausality is a fundamental notion in science and engineering. In the past few decades, some of the most influential developments in the study of causal discovery, causal inference, and the causal treatment of machine learning have resulted from cross-disciplinary efforts. In particular, a number of machine learning and statistical analysis techniques have been developed to tackle classical causal discovery and inference problems. On the other hand, the causal view has been shown to facilitate formulating, understanding, and tackling a broad range of problems, including domain generalization, robustness, trustworthiness, and fairness across machine learning, reinforcement learning, and statistics.\n
\nWe invite papers that describe new theory, methodology and/or applications relevant to any aspect of causal learning and reasoning in the fields of artificial intelligence and statistics. Submitted papers will be evaluated based on their novelty, technical quality, and potential impact. Experimental methods and results are expected to be reproducible, and authors are strongly encouraged to make code and data available. We also encourage submissions of proof-of-concept research that puts forward novel ideas and demonstrates potential for addressing problems at the intersection of causality and machine learning.\n Paper Submission\n The proceedings track is the standard CLeaR paper submission track. Papers will be selected via a rigorous double-blind peer-review process. All accepted papers will be presented at the Conference as contributed talks or as posters and will be published in the Proceedings.\n
\n\nTopics of submission may include, but are not limited to:
\nCLeaR 2025 is being planned as an in-person conference with hybrid elements accommodating online presentations when physical attendance is not possible.
\n\nSubmissions are limited to 12 single-column PMLR-formatted pages, plus unlimited additional pages for references and appendices. Authors of accepted papers will have the option of opting out of the proceedings in favor of a 1-page extended abstract, which will point to an open access archival version of the full paper reviewed for CLeaR. You can also submit a single file of additional supplementary material separately, which may be either a pdf file (containing proof details, for instance) or a zip file that can include multiple files of all formats (such as code or videos). Note that reviewers are under no obligation to examine the supplementary material.
\n\nPlease format the paper using the official LaTeX style files. We do not support submission in formats other than LaTeX. Please do not modify the layout given by the style file.
\n\n\nThe CLeaR review process is double-blind: reviewers and authors will both stay anonymous to each other during the review process. We use OpenReview to host papers; however, public discussions are not allowed during the review process. The review comments are only visible to program chairs, area chairs, and reviewers with submitted review comments. Papers will be desk-rejected if they contain any information that can violate the double-blind reviewing policy, such as the author names or their affiliations, acknowledgements, or links that can infer any author’s identity or institution. Self-citations are allowed as long as anonymity is preserved. It is up to the author’s discretion how best to preserve anonymity when including self-citations. Possibilities include: leaving out a self-citation, including it but replacing the citation text with “removed for anonymous submission,” or leaving the citation as-is. We recommend leaving in a moderate number of self-citations for published or otherwise well-known work.
\n\nRevisions are allowed in the submission system until the paper submission deadline. Changes will not be allowed afterwards.
\n\nWe strongly discourage advertising the preprint on social media or in the press while under submission to CLeaR. Preprints must not be explicitly identified as an CLeaR submission at any time during the review period (i.e., from the abstract submission deadline until the notification of the accept/reject decision).
\n\nCLeaR does not allow double submissions. Namely, submissions should not have been previously published in or submitted to a journal or the proceedings of another conference at any point during the CLeaR review process. Submissions to workshops or other non-archival venues (without a proceedings) will not be considered as dual submissions. Submissions as extended abstracts with 5 pages or less will not be considered a concurrent submission either. Authors may submit anonymized work to CLeaR that is already available as a preprint (e.g., on arXiv) without citing it. Tech reports (including reports on sites such as arXiv) do not count as prior publication. It is acceptable to have a substantially extended version of the submitted paper under consideration simultaneously for journal publication, so long as the journal version’s planned publication date is after our publication (July 15, 2025, tentatively), and it does not violate the journal's policy, the journal submission does not interfere with CLeaR right to publish the paper, and the situation is clearly described at the time of CLeaR submission. Please describe the situation in the appropriate box on the submission page (and do not include author information in the submission itself, to avoid accidental unblinding). Authors are also allowed to give talks to restricted audiences on the work(s) submitted to CLeaR during the review.
\n\nAll accepted papers will be presented at the Conference either as contributed talks or as posters, and will be published in the CLeaR Conference Proceedings in the Journal of Machine Learning Research Workshop and Conference Proceedings series. Papers for talks and posters will be treated equally in publication.
\n\nThe reviewers and area-chairs will have access to papers and supplementary materials that are assigned to them.
\n\nThe program chairs and workflow chairs will have access to all the papers. Everyone having access to papers and supplementary materials will be instructed to keep them confidential during the review process and delete them after the final decisions.
\n\nReviews will be visible to area chairs, program chairs, and workflow chairs throughout the process. At any stage of the process, author names will not be known to the reviewers or area chairs, but only visible to program chairs. Reviewer names are visible to the area chair (and program chairs), but the reviewers will not know names of other reviewers.
\n\nMathias Drton & Biwei Huang
\n CLeaR 2025 Program Chairs
Negar Kiyavash & Jin Tian
\n CLeaR 2025 General Chairs
Thank you for serving as an area chair for CLeaR 2025. As an area chair, your job is to (1) ensure that all the submissions you are assigned have high-quality reviews and good discussions, and (2) write quality meta-reviews and make acceptance decisions. If there is any question, please contact the program chairs at chairs.clear2025@gmail.com. The main tasks of area chairs are listed below.
\n* Please review the policies in the CLeaR 2025 Call for Papers and Reviewer Instructions.
\nAt CLeaR 2025, our goal is to create a community and environment that recognizes and respects the intrinsic value of everyone, which is essential for the open exchange of ideas, freedom of thought and expression, and respectful scientific debate at the conference.
\nAll participants, including but not limited to organizers, reviewers, speakers, sponsors, and volunteers at our Conference and Conference-sponsored social events are required to agree with this Code of Conduct both during an event and on official communication channels, including social media. In particular, sponsors should not use sexual, racial, or other offensive images, events, or other materials. This code applies to both official sponsors and any organization that uses the conference name as a brand as part of its activities during or around the conference.
\nCLeaR is committed to providing all participants with an experience free from harassment, bullying, discrimination and retaliation. This includes offensive comments related to age, gender, gender identity and expression, race, sexual orientation, physical or intellectual disability, physical appearance, body type, ethnicity, religion, politics, technical choices, or any other personal characteristics. No tolerance for bullying, intimidation, personal assault, harassment, continuous interruption of conversations or other activities, and behavior that interferes with the full participation of other participants. This includes sexual harassment, stalking, stalking, harassing photography or recording, inappropriate physical contact, unwelcome sexual attention, vulgar communication, and diminutive characteristics. The policies apply to actual meeting sites and conference venues, including both physical venues, online venues, and official virtual platforms, including but not limited to Openreview comments, video, virtual streaming, Q&A tools. For example, offensive or even threatening comments in Open review are prohibited. Also, Zoom bombing or any virtual activities that have nothing to do with the topic of discussion and that are detrimental to the purpose of the topic or program are not allowed.
\nParticipants who are asked by any member of the community to stop any improper behavior defined here should immediately comply. Meeting organizers may take further actions at their discretion, including: formally or informally warning offenders, expelling the meeting without refunds, refusing to submit or participate in future CLeaR meetings, reporting the incident to the offender’s local agency or funding agency, or reporting to the local Authorities or law enforcement agencies report the incident. The ''just kidding'' response is unacceptable. If action is taken, an appeal procedure will be provided.
\nIf you have any concerns about possible violation of the policies, please contact the conference chairs (chairs.clear2025@gmail.com) as soon as possible. Reports made during the conference will be responded to within 24 hours; those at other times in less than two weeks. Complaints and violations will be handled with discretion. We are ready and eager to help attendees contact relevant help services, escort them to a safe place, or to otherwise help people who have been harassed feel safe during the meeting. We gratefully accept feedback from the CLeaR community on our policy and actions.
\nA similar version has been used by other conferences, such as ICLR 2020 and UAI 2020.
\nThis page defines what constitutes a conflict of interest for the CLeaR 2025 review process.
\nIf you are an author, reviewer, or area chair, please make sure to create or update your OpenReview profile. You will be asked to declare two types of conflicts---domain conflicts and personal conflicts. Both types are declared by filling out appropriate sections of your OpenReview profile, as described below.
\nTBD
\n\nCLeaR 2025 will be held at the SwissTech Convention Center, Lausanne, Switzerland, on May 7-9, 2025.\n
\nYou can register for CLeaR 2025 through this form. As we have a limited number of available tickets, please register as soon as possible. Please indicate whether you need an invitation letter for a visa application (at the bottom of the registration form).
\nBelow please find preferential rates for accommodations.
\nTo reserve a room: send an email to reception@sthotel.ch, with the following information:
\nReservation code: CLEAR25
\nRoom price: CHF 130.--
\nBreakfast: CHF 16.--/person
\nTourist taxes: CHF 5.50/person
\nTo reserve a room before April 6, 2025: http://bookings.ihotelier.com/bookings.jsp?groupID=4475883&hotelID=73897
\nTo reserve a room from April 7, 2025 on: send an email to s.mesnil@shlausanne.ch with code CLEAR25, groupID 4475883
\nRoom price per person: CHF 170.--
\nSupplement for double occupancy: CHF 35.--
\nBreakfast buffet and parking included
\nTourist taxes: CHF 6.--/person
\nTo reserve a room: send an email to booking@hotelcontinental.ch, with the following information:
\nReservation code: CLEAR25
\nRoom price: CHF 195.--
\nBreakfast: included
\nBus and metro tickets: included
\nTourist taxes: CHF 6.--/person
\nBaggage taxes: CHF 4.--/baggage (if needed)
\nTo reserve a room: use this link to book with group rate for EPFL until March 26, 2025 (after this date, the price will be the regular market price). At your arrival, you will have to deposit a credit card number for guarantee.
\nRoom price single room: CHF 190.--
\nRoom price double room: CHF 205.--
\nBreakfast: included
\nTourist taxes: CHF 5.50/person
\nWe highly recommend to attend the conference in-person, even more so for presenting your accepted papers.
\nPlease contact us at zietld@amazon.com in case you cannot present (oral or poster) in-person (subject: [CleaR 2025 remote]).
\nTime: {time}
\nJoin Zoom Meeting:
\n \nMeeting ID: {meetingId}
\nPasscode: {passcode}
\n \nRegistered participants can attend keynotes and oral presentations (see schedule) remotely. Please connect to this Zoom meeting:
\nThis link is used for all keynotes and orals. Note that in-person poster sessions, the townhall, panel discussion, and social activities will not be streamed.
\nWhen participating remotely, please mute yourself (except when asking questions). Feel free to ask questions either in the chat or by raising your hand in Zoom and asking them over video.
\nVirtual posters can be found below.
\n| Paper title | \nDownload link | \n
|---|---|
| {paper.title} | \nDownload | \n
TBD
\n