call for papers
Important dates
All times are end of day AOE.
- submission deadline: 22nd Nov 2024
- expected notification of acceptance: 9th Dec 2024
- camera-ready deadline: 10th Jan 2025
- workshop: 4th March 2025
Topics of interest
Topics of interest include, but are not limited to:
-
low-rank structures to speed up computation of large ML systems such as adaptors in LLMs and structured computational graphs
-
circuit representations for reliable and efficient probabilistic reasoning and learning with applications to trustworthy ML such as reliable neuro-symbolic AI
-
tensorial architecture design in deep learning such as polynomial representations that have emerged as a strong-performing alternative to the standard neural network paradigm
-
tensor networks for quantum and physics-inspired computing to solve variational inference, PDEs and inverse problems
-
theory of matrix and tensor factorization methods and their optimization in sketching, compression and tensor completion
-
expressivity of tensor representations including exponential lower bounds of matrix ranks and circuit sizes and sample complexity of learned representations
-
applications of low-rank representations in AI-related fields such as data and model compression, multimodal fusion, and model interpretability.
Submissions
We invite two types of submissions:
- novel research on low-rank representations in AI or retrospective papers on lessons learned and reproducibility studies (up to 4 pages, excluding references in AAAI format)
- papers already accepted at a major AI and ML conference this year (in the original format and length, not anonymized)
Submissions will be accepted only through OpenReview and must closely follow the formatting guidelines in the templates, otherwise, they will be desk-rejected.
Supplementary material can be put in the same pdf paper (after references); it is entirely up to the reviewers to decide whether they wish to consult this additional material.
Reviewing
Reviewing for novel research papers is double-blind; i.e., reviewers will not know the authors’ identity and authors won’t know the reviewers’ identity. Please refer to your prior work in the third person wherever possible. We also encourage links to public repositories such as GitHub to share code and/or data as long as they are anonymized. Reviewing for already accepted papers is single-blind: authors won’t know the reviewers’ identity.
List of reviewers:
TBD