MathAtlas is the first large-scale benchmark for autoformalizing graduate mathematics, where even strong models reach only 9.8% correctness on theorem statements and drop to 2.6% on the hardest dependency-deep subset.
emnlp-main.233/
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
years
2026 2representative citing papers
A monotonic reference-free iterative process for full-theorem autoformalization simultaneously improves formal validity, logical preservation, mathematical consistency, and formal quality, reaching 100% formal validity on miniF2F.
citing papers explorer
-
MathAtlas: A Benchmark for Autoformalization in the Wild
MathAtlas is the first large-scale benchmark for autoformalizing graduate mathematics, where even strong models reach only 9.8% correctness on theorem statements and drop to 2.6% on the hardest dependency-deep subset.
-
Monotonic Reference-Free Refinement for Autoformalization
A monotonic reference-free iterative process for full-theorem autoformalization simultaneously improves formal validity, logical preservation, mathematical consistency, and formal quality, reaching 100% formal validity on miniF2F.