Today for AI
HOT RADAR
llmHEAT 11.9°

AI CURATED HOT STORY · 11/2/2026

Meta and NYU Study Shows Simple Loop with Lean Tools Beats Complex Theorem Proving Workflows

1 reports archived1 independent sourcesupdated 11/2/2026, 03:25:28
Synthesis & Latest Updates
1 Sources Cross-Validated

A joint paper by Meta and NYU introduces 'h1', a simple loop system that combines Lean formal verification feedback with search tools. It outperforms the hand-designed Goedel-Architect workflow at lower cost using the same model, demonstrating that effective tool interfaces and compute scaling are more cost-efficient than complex planning logic in theorem proving.

LATEST/A joint paper by Meta and NYU introduces 'h1', a simple loop system that combines Lean formal verification feedback with search tools. It outperforms the hand-designed Goedel-Architect workflow at lower cost using the same model, demonstrating that effective tool interfaces and compute scaling are more cost-efficient than complex planning logic in theorem proving.

Hourly heat curve

2 fully observed hours
Fewer than 3 fully observed hours; trend not plotted yet.

Coverage timeline

Total 1 reports · Latest first
  1. Rohan Paul (@rohanpaul_ai)T2·78 pts

    Meta and NYU Study Shows Simple Loop with Lean Tools Beats Complex Theorem Proving Workflows

    • The h1 system solved 598 problems on PutnamBench at $1.08/problem, outperforming Goedel-Architect's 597 problems at $1.47/problem.
    • Core mechanism relies on fast Lean feedback and search tools, maintaining progress summaries after failures instead of pre-planning.
    • Interface design is critical: With Opus 4.8, h1 solved 83% of hard algebra problems compared to 63% for Claude Code.