38 min
Seminar, Hosted by Harvard Center of Mathematical Sciences and Applications — New Technologies in Mathematics Seminar
The IEANTN Project: Automation and Mathematical Understanding
UCLA
Recording
Abstract
Explicit analytic number theory combines conceptual arguments with heavy bookkeeping and numerical computation, often relying on long chains of results spread across many papers. The IEANTN project organises this literature into an evolving network that makes dependencies and supporting evidence explicit. Xinjie He explains how the project was redesigned: version 1.0 kept everything in a single Lean repository, which became hard to maintain and rebuild as auto-formalised proofs accumulated, while version 2.0 separates human-readable statements from verified proofs and records different levels of justification. The talk weighs which parts of the work should be handed to automation and which mathematicians want to understand themselves, arguing that a network view can make the ideas and their connections clearer while machines handle the detailed bookkeeping.