- Call for Papers released:
Dec. 16th, 2022
- Paper submission deadline:
Feb. 8th, 2023 (AoE)
- Review period starts:
Feb. 11th, 2023
- Review period ends: Feb. 25th, 2023
- Reviews released: Mar. 3rd, 2023
- Camera-ready deadline: Apr. 20th, 2023 (AoE)
- Hybrid workshop date: May 4th, 2023
Neurosymbolic generative models have recently gained significant attention as a promising approach to integrate the strengths of both symbolic and neural models. These models combine the ability of symbolic models to explicitly represent abstract concepts and logical reasoning with the ability of neural models to learn and generalize from data.
We invite researchers to submit their papers, both theoretical and empirical, on all aspects of Neurosymbolic generative models, including but explicitly not limited to:
- Neurosymbolic Programming
- Generative models for discrete structures (e.g., generative models of graphs)
- Generative models with symbolic background knowledge
- Probabilistic circuits and tractable models
- Neurosymbolic reasoning
- Causality-inspired generative models
- Semantic Parsing
- Structured world models
- Object-centric learning
- Evaluation and comparison of neurosymbolic generative models to regular generative models
- Applications of Neurosymbolic generative modeling (e.g., natural language processing, computer vision, programming, robotics)
We invite both new and unpublished short papers, and extended abstracts of recently published work, with a maximum of 4 pages (excluding references and appendices). We will do light double-blind reviewing. Reviews will remain closed after the decisions, and this workshop will have no proceedings (it is non-archival). Dual submission is allowed where permitted by third parties.
NeSy-GeMs will be a hybrid workshop: Authors can present a poster in person if they come to Kigali. If authors can not attend physically, they can present their posters virtually. To prepare your submission, please use the LaTeX style files for ICLR 2023, provided at https://github.com/ICLR/Master-Template/raw/master/iclr2023.zip. The papers should be prepared for double-blind review.
Please ask on our Google Groups or mail us when in doubt if a paper fits the topic of the workshop. Researchers in the broader fields of generative modeling and Neurosymbolic AI are encouraged to submit papers if the relevance to the topic of the workshop is clear.
Submit your papers on Openreview.
Will the OpenReview be public? Submissions will be private until the author notification (Mar 3rd, 2023). Rejected and withdrawn submissions will not be made public.
What is the dual submission policy? We accept
- Original work not submitted anywhere else
- Extended abstract of work published in other journals or conferences (including ICLR 2023)
Are appendices allowed? Yes, they are. They will not count towards the page count (and neither will references).