Workshop on Associative Memory @ICLR 2026

Submit Your Paper

Camera ready due Mar 20.

Submissions are now closed.

5-page full papers or 3-page tiny papers (excluding references and supplementary). Non-archival, double-blind review.

Submit on OpenReview LaTeX Template Call for Papers

Camera-ready instructions

Submit your deanonymized paper on OpenReview. The final PDF should follow the official NFAM workshop ICLR format.

If you are using the LaTeX template, add \iclrfinalcopy before \begin{document}. Include author names and institutions in the final version.

Page limits stay the same: 3 pages for tiny papers and 5 pages for full papers. Acknowledgements and references do not count toward the limit.

Submission Feb 14

Notification Mar 1

Camera Ready Mar 20

Workshop Apr 26

About the Workshop

Associative Memory has re-emerged as a unifying principle linking classical energy-based models with transformers, diffusion models, and memory-augmented agents. This workshop brings together theorists and practitioners to develop a unified perspective on memory as the core substrate of intelligent behavior. This is the third iteration of the workshop, following ICLR 2025 and NeurIPS 2023.

Read more about the workshop scope

Rooted in early mathematical and neuroscientific formulations, associative memory (AM) provides a principled view of collective computation through attractor dynamics and energy landscapes. Recent advances have significantly expanded this framework, revealing associative retrieval as a form of attention, inference, and optimization within deep learning systems. Beyond classical recall, modern AM models support test-time regression, continual adaptation, and reasoning over structured domains such as graphs, manifolds, and probability distributions.

Despite rapid progress, research on memory, reasoning, and adaptation remains fragmented across communities spanning energy-based learning, optimization theory, neuroscience-inspired computation, generative modeling, and agentic AI. By fostering cross-disciplinary dialogue and shared evaluation paradigms, the workshop aims to catalyze a coherent research agenda for next-generation memory-augmented and agentic AI systems.

Submission details

We invite submissions on novel research results (theoretical and empirical), benchmarks, demos, visualizations, software frameworks, and work-in-progress research. Submissions should be made anonymously on OpenReview using this workshop's LaTeX template. Reviews will not be shared publicly. See the call for papers for more details.

Supplementary materials after references are allowed and don't count toward the page limit. Submissions should be anonymized. Accepted papers will be made public on OpenReview. The reviewing process is double-blind with standard ICLR conflict of interest rules. Preference given to new, unpublished work.