Development

FrontierMath Tier 4 jumps from 19% to 31%

development2026-01-25

Claim

GPT-5.2 Pro scored a state-of-the-art 31% on FrontierMath Tier 4, up from 19%, with mathematicians now studying the cases where it failed unexpectedly.

Reported metric: 31%

Reporting date: 2026-01-25 (not necessarily the event date)

Sources

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

Welcome to January 25, 2026 (theinnermostloop.substack.com)

The interesting work moves to the failures.

Knowledge graphopen in browser →

About

FrontierMath

Reported in

Welcome to January 25, 2026

Actor

OpenAI

Evidence for

Supersedes

GPT-5.2 Pro takes 29.2% on FrontierMath Tier 4

Referenced by

Cite or correct this record

Citation formatCitation text

FrontierMath Tier 4 jumps from 19% to 31%
Innermost Loop KB — secondary record
Reporting date: 2026-01-25
KB permalink: https://nicholsn.github.io/innermost-loop-kb/developments/2026-01-25-frontiermath-tier-4-31pct
Claim: GPT-5.2 Pro scored a state-of-the-art 31% on FrontierMath Tier 4, up from 19%, with mathematicians now studying the cases where it failed unexpectedly.
Attributed sources:
- Welcome to January 25, 2026 — https://theinnermostloop.substack.com/p/welcome-to-january-25-2026
This citation does not imply independent verification of linked-source claims.

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