Call for Papers

We invite researchers working on theoretical aspects of foundation models to submit their work for consideration in the TF2M workshop. We welcome submissions that make theoretical contributions on topics including, but not limited to:

• Efficient training, finetuning, and inference algorithms.

• Data-efficient training and fine-tuning strategies.

• Theoretical foundations of model compression, pruning, and distillation.

• Fairness and bias mitigation in foundation models.

• Principles of model alignment, and safety.

• Directions in privacy and security for foundation models.

• Statistical and information-theoretic perspectives on model capabilities.

• Optimization theory for model training and fine-tuning.

• Emergent capabilities of LLMs, such as in-context learning.

• Understanding of neural architectures behind modern neural models such as transformers.

Submission Details

Authors should upload a short paper of up to four pages in ICML format (for camera-ready, please using the updated template), with unlimited pages of references and supplementary material. Note that the reviewers are not required to read the appendices and the claims of the paper should be supported by material in the main four page body. Please submit a single PDF that includes the main paper and supplementary material. We welcome submissions that present work which is unpublished or currently under submission. We will also consider recently published (i.e., in 2024) work in venues other than ICML 2024 and ICLR 2024.

All submissions will be reviewed in a double-blind process and will be evaluated on the basis of their technical content and relevance to the workshop. Accepted papers will be selected to be presented either in a poster session or as a contributed talk. This workshop is non-archival and submissions can be submitted to other venues. The accepted papers will be publicly available through openreview before the start of the workshop.

If you have any questions, please contact us at tf2m.icml2024@gmail.com.

Submission link: OpenReview.

Important Dates

Full Paper Submission Deadline: May 29th, 2024, May 31st, 2024,  11:59 pm AOE

Accept/Reject Notification Date: June 17th, 2024, 11:59 pm AOE

Camera-ready Submission Deadline: July 20th, 2024, 11:59 pm AOE

Workshop Date: July 27th, 2024 (Vienna, Austria)

Volunteering as a Reviewer

If you would like to volunteer as a reviewer, please fill out this form.