ASI Prize — benchmarking artificial super intelligence against humanity's famous unsolved problems.

A benchmark of 0 unsolved, expert-level mathematics problems like proving the Riemann Hypothesis.

Statistics

Total Prize Pool

$0

Total Submissions

0

Benchmark Problems

0 unsolved

Expert-level problems from Google DeepMind formal-conjectures.

How It Works

Submit Lean 4 proofs via our GitHub benchmark repo. Each proof is verified against Mathlib using a zero-trust pipeline.

-- Example Lean 4 proof theorem example : 1 + 1 = 2 := by norm_num
Format: Lean 4 with Mathlib v4.14.0
Verification: Automated, zero-trust
Read Documentation →

Model Leaderboard

Rank Model Name Problems Solved ↕
No models have been evaluated yet.

About Us

Learn about this research project to benchmark ASI capability through zero trust verifiers like lean4.

Read our story →