Full papers
Up to 4 pages (excluding references, acknowledgments, or appendices), submitted via OpenReview, with double-blind review; accepted papers receive either a poster or talk slot.
Machine Learning/CORE Unranked
ICML MemFM 2026
The Second Workshop on the Impact of Memorization on Trustworthy Foundation Models at ICML
The MemFM 2026 workshop, held in conjunction with ICML 2026 in Seoul, focuses on the critical issue of unintended memorization in foundation models and its impact on trustworthiness, privacy, and ethical AI. It brings together researchers across machine learning security, data privacy, and AI ethics to explore detection, mitigation, theoretical foundations, and societal implications of memorization in high-stakes applications.
Paper fit
Contribution paths
A strong submission should clearly identify its contribution and evaluate it appropriately.
Research areas in scope
01
Research Areas
Detection and Mitigation Methods for Memorization in Foundation ModelsTheoretical Foundations of MemorizationQuantification of the Degree of MemorizationRelationships Between Memorization and Security, Privacy, and SafetyImplications of Memorization on Generalization in Foundation ModelsConnecting Memorization Research Across Different Domains and ApplicationsSocietal Impact and Ethical Aspects of MemorizationLegal Perspectives on Memorization and Intellectual Property
Policies worth checking twice
- Submissions must follow ICML double-blind reviewing guidelines.
- Papers must be anonymous and comply with ICML Code of Conduct and Code of Ethics.
- Dual submissions with other workshops or conferences are allowed.
- The workshop is non-archival and does not have official proceedings.
- Accepted papers can be submitted elsewhere if permitted by their policies.
- The 4-page limit applies to both initial and camera-ready submissions.
- Each submission must have at least one author serving as a reviewer; failure results in desk rejection.
- Best Paper Award will be given to one accepted paper.
Official sources
Compiled from the official call for papers. The organizers’ pages remain authoritative.
Last verified September 9, 2026