← back to claim Held-out contamination audit found miniF2F-Lean4 gains persist at 54% on 300 freshly authored problems never posted online

evidence graph

depth 1 2 3 both pro con
cited by (1)
[+0] Autoformalization success on undergraduate competition corpora rose from 8% to 61% between 2022 and 2025 on the miniF2F-Lean4 benchmark [i] formal-verification machine-learning
Held-out contamination audit found miniF2F-Lean4 gains persist at 54% on 300 freshly authored problems never posted online
[+0] validity 0.0 centrality 22.5 consensus 0 depth 0.0
formal-verification machine-learning virtues
opposes (0)
no opposing evidence yet
whole evidence graph shown
inspector

Select a link’s marker to see who asserted it, why, and what it cites.