Development

A fresh olympiad benchmark more than doubles in a single release

development2026-04-29

Claim

GPT-5.5 scored a record 73.66% on fresh olympiad problems, more than doubling its predecessor's 36.61% and repricing what counts as a hard problem.

Reported metric: 36.61% → 73.66%

Reporting date: 2026-04-29 (not necessarily the event date)

Sources

Attribution and reporting linked by this record; links do not imply independent verification.

Welcome to April 29, 2026 (theinnermostloop.substack.com)

Knowledge graphopen in browser →

Reported in

Welcome to April 29, 2026

Actor

OpenAI

Evidence for

Supersedes

A 23-year-old with no advanced training solves an Erdős problem in one prompt

Referenced by

Cite or correct this record

Citation formatCitation text

A fresh olympiad benchmark more than doubles in a single release
Innermost Loop KB — secondary record
Reporting date: 2026-04-29
KB permalink: https://nicholsn.github.io/innermost-loop-kb/developments/2026-04-29-matharena-doubles-in-one-release
Claim: GPT-5.5 scored a record 73.66% on fresh olympiad problems, more than doubling its predecessor's 36.61% and repricing what counts as a hard problem.
Attributed sources:
- Welcome to April 29, 2026 — https://theinnermostloop.substack.com/p/welcome-to-april-29-2026
This citation does not imply independent verification of linked-source claims.

Suggest a correction — opens a GitHub draft for you to review and submit.