Model comparison

$ ./verieval-dashboard

Leaderboard Compare Heatmap Errors Dataset Proofs Runs

Verifier Dataset Subset Task Mode k Sweep From To show excluded

Leaderboard group by model show hidden

y x log x labels families arrow

Difficulty heatmap

Solved Not solved Not attempted

Error categorization — all models

Dataset & problem statistics

Proof length by model

Runs & progress show hidden local

Per-model breakdown

Problem	Solved	Samples — click to inspect

Sample trace

Task