Submissions Open! Mechanistic Interpretability Workshop at NeurIPS 2025
Submissions are now open for the Mechanistic Interpretability Workshop at NeurIPS 2025! The submission deadline is August 22nd, and you can submit papers on OpenReview here. Papers under review for NeurIPS are welcome.
We welcome all papers which further the field of mechanistic interpretability, i.e. which advance our ability to use the internal states of neural networks to understand them. We accept both short (4 page) and long (9 page) papers, and we take an expansive view of what furthering the field can mean! See our call for papers for more information.
We also need reviewers. If you've written a mechanistic interpretability paper, we'd love your help. Please express interest to be a reviewer here.
Looking forward to your contributions,
The Mech Interp Workshop Organizing Committee