[KongchangAI]
Benchmark

Putnam Bench

A benchmark test set for evaluating automated theorem proving systems, containing a large number of formalized proof problems of the mathematical competition type.

Timeline (last 90 days)

Nov 17

Under matching the same model and comparable budgets, H1 matches or Pareto-dominates custom systems on multiple benchmarks including Putnam Bench, and has solved eight open conjecture problems.

Unverified50%
Nov 17

On Putnam Bench, when using the same model as Goedel Architect, H1 is cheaper but performs better.

Unverified50%
Nov 17

When using GPT 5.6 Sol, H1 achieves a 90-100 solve rate on FateX, 178% on FC100 Soft, 23-14% on Formal Core Bench, solves all 6 problems of IMO2026, and solves 670 of the 672 problems in Putnam Bench.

Unverified50%

All Facts (3)

Source Articles