Interpretable AI Workshop

Calls

    Call for Papers

    We invite submissions to the workshop on interpretability in machine learning, titled “Interpretable AI: Past, Present and Future”. This workshop aims to bridge classical interpretability research with modern methods for foundation models, fostering discussions on the mathematical foundations, methodology, design, evaluation, and application of interpretable machine learning models.

    Topics of Interest

    We welcome submissions addressing, but not limited to, the following topics:

    • Novel approaches, methodologies, and foundations for interpretability in machine learning
    • Metrics and benchmarks for evaluating the quality and reliability of interpretable models
    • Mechanistic interpretability methods for foundation models (e.g.: Large Language Models and Diffusion models)
    • Comparative studies on different interpretability methods and their applications
    • Integration of domain knowledge and expertise in interpretable model design
    • Applications of interpretability in diverse domains such as healthcare, earth sciences, material sciences, physics, and beyond
    • Exploration of ethical and legal needs for enforcing interpretable models in decision-making processes

    Submission Guidelines

    Authors are invited to submit up to 6 pages with unlimited references and supplementary materials. Submissions should be anonymized and should use the official template. Please submit your papers through the official openreview portal.

    We have three tracks: main track, published paper track, and position paper track.

    Reviewing is double-blind. Submissions should clearly articulate the research problem, methodology, results (if available), and contributions to the field.

    For the main track, submissions can be exploratory in nature and contain initial novel ideas and results. We welcome submissions undergoing concurrent peer review elsewhere. Papers presented or scheduled for presentation at non-archival venues, such as other workshops, are permitted for submission. However, it is the authors’ responsibility to verify compliance with other venues’ policies.

    Submissions that have been previously published or accepted for publication in peer-reviewed conferences or journals should be submitted to the published paper track. These papers will not undergo the review process and will instead be evaluated by ACs based on the paper’s topic and its fit for the workshop.

    We also invite position papers that provide a commentary on the current state of interpretable machine learning or its usage in practice, as these may be of broader interest. To clearly identify submissions, position papers must be titled beginning with the phrase “Position: <Paper title>” and submitted to the position paper track.

    This workshop is non-archival, and will not have formal proceedings. All the accepted papers will be publicly available on our workshop website. The selection of oral presentations will be made based on the PC’s recommendations and their fit to the topics of the workshop.

    Important Dates

    • Submission open on OpenReview: August 9, 2024
    • Submission Deadline: August 30, 2024
    • Notification of Acceptance: October 9, 2024
    • Camera-ready Deadline: November 15, 2024
    • Workshop Date: December 15, 2024

    All deadlines are 11:59PM UTC-12:00 (AoE).

    Contact

    For inquiries, please contact interpretable.ai.neurips.workshop [AT] gmail.com.