Vibeleaderboard
← All Intel
Intel / article

The Problem Is the Problem: Towards Scalable Mathematical Discovery

Source
arxiv.org
Author
Zeyu Zheng, Shengtong Zhang, Jeremy Avigad, Prasad Tetali, Sean Welleck
Date

Context

Find, Attempt, and Recommend changes where mathematicians enter an AI research workflow. Experts choose a research direction; the system searches papers for candidate problems, filters them, attempts solutions and recommends selected results for review. The aim is to spend scarce mathematical on promising work. The paper calls this pipeline FAR.

In a combinatorics pilot starting with 5,245 papers, the pipeline surfaced 598 potential resolutions and selected 77 items for the author team to review. Those counts describe successive filters, not a tally of verified discoveries. The contribution is a method for allocating problem-solving and review effort across a literature-derived pool.

Terms in this piece · Glossary
  • attention — The mechanism that lets a model weigh which earlier words matter for the word it's currently processing — the core operation of a transformer.
Recommended reads
Comments

Checking sign-in…

Loading comments…