anthropics/math-olympiad
claude-codeOfficial# math-olympiad Competition math solver with adversarial verification. ## The problem Self-verification gets fooled. A verifier that sees the reasoning is biased toward agreement. ArXiv:2503.21934 ("Proof or Bluff") showed 85.7% self-verified IMO success drops to <5% under human grading. ## The approach - **Context-isolated verification**: verifier sees only the clean proof, never the reasoning trace - **Pattern-armed adversarial checks**: not "is this correct?" but "does this accidenta
Our Verdict
Competition math solver with adversarial verification for olympiad-level problems. Suited for AI researchers improving math reasoning models. Stands out with context-isolated checks: verifier views only clean proofs, dodging bias from reasoning traces.
Frequently Asked Questions
What is anthropics/math-olympiad used for?
It solves competition math problems using adversarial verification to counter self-verification bias. The verifier sees only the clean proof without reasoning traces, addressing issues where 85.7% self-verified IMO success drops below 5% under human grading per ArXiv:2503.21934.
What is math-olympiad?
# math-olympiad Competition math solver with adversarial verification. ## The problem Self-verification gets fooled. A verifier that sees the reasoning is biased toward agreement. ArXiv:2503.21934 ("Proof or Bluff") showed 85.7% self-verified IMO success drops to <5% under human grading. ## The approach - **Context-isolated verification**: verifier sees only the clean proof, never the reasoning trace - **Pattern-armed adversarial checks**: not "is this correct?" but "does this accidenta
How do I install math-olympiad?
Visit the GitHub repository at https://github.com/anthropics/claude-plugins-official/tree/main/plugins/math-olympiad for installation instructions.
What license does math-olympiad use?
math-olympiad uses the Proprietary license.
What are alternatives to math-olympiad?
Explore related tools and alternatives on My AI Guide.
Open source: source code publicly visible
Anyone can inspect exactly what this repo does on GitHub before using it. Proprietary licensed.
Reviewed by My AI Guide for relevance, quality, and active maintenance before listing.
Install in Claude Code:
/install math-olympiad