We welcome submission on the following non-exhaustive topics, touching on systematic generalization, out-of-sample / out-of-domain generalization, learning from non-IID/long-tail data:

  • Demonstration of success and failure modes in biological intelligence versus neural networks or other models for artificial intelligence.

  • Computational theory for generalization in biological or artificial intelligence.

  • Mechanisms of biological intelligence that support generalization, at the level of cells, circuits, systems, or cognition.

  • Datasets or benchmarks that represent novel challenges for biological or artificial intelligence.

Important Dates (anywhere on earth)

  • Paper submission deadline: Extended to 20 March 2021

  • Acceptance notification: 26 March 2021

  • Workshop: Friday, May 7th, 2021 (Note this new date!)

Author Instructions:

Abstracts can contain up to, but not more than 300 words, and may include a maximum of 1 figure. Submissions may be work in progress or currently under review - as long as the other venue’s rules permit that. The format of the submission should be a pdf with the ICLR template.

Abstracts are submitted via CMT.

Evaluation

Submissions will be evaluated by the workshop organizers. We intend to accept the majority of submissions. Reasons for rejection could be an inappropriate topic, flawed or no methods, incomprehensibility or repetition of known literature.

Accepted abstracts

Accepted abstracts will be presented during the poster session. They will also be posted on the workshop site, but under a non-archival status.