Full papers
Original research contributions with substantial theoretical or empirical results.
Machine Learning/CORE Unranked
HiLD at ICML 2026
High-dimensional Learning Dynamics 2026
The Workshop on High-dimensional Learning Dynamics (HiLD) at ICML 2026 focuses on the theoretical and practical aspects of learning dynamics in high-dimensional spaces, bringing together researchers to explore novel methods, analyses, and applications in machine learning where dimensionality plays a critical role.
Paper fit
Contribution paths
A strong submission should clearly identify its contribution and evaluate it appropriately.
Short papers
Preliminary results, novel ideas, or work in progress with clear potential.
Journal track submissions
Extended versions of previously published conference papers, suitable for journal publication.
Research areas in scope
01
Topics of Interest
Theoretical foundations of high-dimensional learning dynamicsDynamics of optimization in overparameterized modelsGeneralization and implicit regularization in high dimensionsNeural tangent kernels and feature learning dynamicsHigh-dimensional statistical physics of learningDynamics of representation learning and feature emergenceAdversarial robustness and learning dynamicsDynamics of reinforcement learning in high-dimensional state spacesInterplay between architecture, initialization, and learning dynamicsEmpirical studies of learning dynamics in large-scale models
Policies worth checking twice
- Submissions must be anonymized for double-blind review.
- Full papers are limited to 8 pages, short papers to 4 pages, excluding references.
- Dual submission to other venues is not allowed during review.
- Author lists and affiliations may be updated after acceptance.
- Authors must include a statement on the use of AI-assisted tools in the submission.
Official sources
Compiled from the official call for papers. The organizers’ pages remain authoritative.
Last verified September 9, 2026