• Home
  • Technology
  • Gaming
  • Entertainment
  • World & Business
  • Science
  • Sports
  • AI
HomeTechnologyGamingEntertainmentWorld & BusinessScienceSportsAI
AI
Reaction

A proof-quality benchmark for Lean is reportedly in development

A user points to the Desargues team’s work and argues that proving involves taste, style and “interestingness” as well as correctness.

1 Source, 11d ago, first seen 11d ago

TLDR

A user says Desargues is working on a proof-quality benchmark for Lean. They argue that improving agents’ taste, style and “interestingness” in proofs could matter for people’s ability to understand and interact with modern synthetic formal mathematics.

Combined views

—

1 Source, first seen 11d ago

— likes— comments— saves— reposts

Combined views

—

1 Source, first seen 11d ago

— likes— comments— saves— reposts

Sentiment

Positive——Negative

Summary

Not enough discussion yet.

No sentiment analysis available yet.

Featured Source

Sentiment

Positive——Negative

Summary

Not enough discussion yet.

No sentiment analysis available yet.

Related

Desargues gets @spolu's backing; a canonical Lean benchmark is in the works

A post says @spolu backed Desargues and that its author is building a public benchmark for canonical Lean, pointing to minif2f as an example of what such a benchmark can do for a field.

Code contracts as a proposed middle ground between informal specs and formal verification

In a DotAI 2026 talk, a developer argues that code remains essential to understanding and maintaining critical business systems, even if people no longer write it themselves.

1 Source

Stanislas Polu@spoluPretty cool work from http://desargues.ai team working on a proof quality benchmark for Lean https://www.desargues.ai/preview/formal-code-quality Proving is not just about correctness it’s also about taste, style, and “interestingness”. Pushing agents to be better there seems like a pretty important thing if we want to maintain our capability to understand and interact with modern synthetic formal mathematics.11d
    • Home
    • Technology
    • Gaming
    • Entertainment
    • World & Business
    • Science
    • Sports
    • AI
    LeanStanislas Polu

    1 Source

    Stanislas Polu@spoluPretty cool work from http://desargues.ai team working on a proof quality benchmark for Lean https://www.desargues.ai/preview/formal-code-quality Proving is not just about correctness it’s also about taste, style, and “interestingness”. Pushing agents to be better there seems like a pretty important thing if we want to maintain our capability to understand and interact with modern synthetic formal mathematics.11d
    Today's Rank

    —

    Not ranked yet

    Today's Rank

    —

    Not ranked yet