Call for Papers

The Math-AI workshop is intended to provide a forum for discussing missing elements and major bottlenecks towards demonstrating mathematical reasoning ability in AI systems. We hope that the outcome of the workshop will lead us in meaningful directions towards a generic approach to mathematical reasoning, and shed light on general reasoning mechanisms for artificial intelligence.

In particular, we are but not limited to the following areas of questions:

Mathematical reasoning vs. general intelligence?

• What is a potential path from mathematical reasoning to general artificial intelligence?
• What is special about mathematical reasoning compared to other reasoning tasks?
• What are the pros and cons of studying mathematical reasoning?
• Can the lessons learned from mathematical reasoning generalize to other reasoning tasks?

What machine learning techniques are we missing in the quest for machines that perform mathematical reasoning?

• How to deal with large action space – hopeless exploration in mathematical reasoning?
• How to design methods that allow flexible planning and jumpy reasoning?
• How to collect large scale datasets for mathematical reasoning?
• Can one augment the dataset by designing good synthetic datasets?
• How do we utilize informal mathematical datasets?
• Human-like theorem proving vs formal theorem proving using ITPs (interactive theorem provers), trade-offs?
• Do we need a special theorem proving assistant for machine learning?

We are also interested in works on other reasoning problems that share similar structures, and can bring great insights to mathematical reasoning, such as,

• Program synthesis
• Code verification
• Retrosynthesis problem
• Text games

Important Dates

Submission: Feb 26 (11:59pm Pacific Time)
Notification: March 26 (11:59pm Pacific Time)
Camera Ready: April 12 (11:59pm A.O.E.)
Workshop: May 07

Submission Requirements

Submissions to Math-AI 2021 are limited to 4 pages of content, but may contain an unlimited number of pages for references and appendices. The latter may not necessarily be read by the reviewers. We request and recommend that authors rely on the supplementary material only to include minor details (e.g., hyperparameter settings, reproducibility information, etc.) that do not fit in the 4 pages. The submission process is double-blind.

All submissions must be formatted with LaTeX using the ICLR paper format(adapted).

All accepted papers will be presented in a virtual poster session, and some will be selected for oral presentation. We welcome articles currently under review or papers planned for publication elsewhere. However, papers that have been published at an ML conference should not be submitted. Accepted papers will be published on the MATH-AI homepage, but are to be considered non-archival.

Submission Link: https://cmt3.research.microsoft.com/MATHAIICLR2021

Please email any enquiries to mathai.iclr2021@gmail.com