Development

Orchestrating three labs' models takes 55% on Humanity's Last Exam

development2026-02-11

Claim

Poetiq reached a state-of-the-art 55% on Humanity's Last Exam by orchestrating Gemini, GPT and Claude together, while Unsloth released Triton kernels enabling twelvefold faster training with 35% less VRAM.

Reported metric: 55% / 12x

Reporting date: 2026-02-11 (not necessarily the event date)

Sources

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

Welcome to February 11, 2026 (theinnermostloop.substack.com)

Knowledge graphopen in browser →

About

Humanity's Last Exam

Reported in

Welcome to February 11, 2026

Actor

Evidence for

Supersedes

Opus 4.6 sets a new high on Humanity's Last Exam

Referenced by

Cite or correct this record

Citation formatCitation text

Orchestrating three labs' models takes 55% on Humanity's Last Exam
Innermost Loop KB — secondary record
Reporting date: 2026-02-11
KB permalink: https://nicholsn.github.io/innermost-loop-kb/developments/2026-02-11-poetiq-55pct-hle
Claim: Poetiq reached a state-of-the-art 55% on Humanity's Last Exam by orchestrating Gemini, GPT and Claude together, while Unsloth released Triton kernels enabling twelvefold faster training with 35% less VRAM.
Attributed sources:
- Welcome to February 11, 2026 — https://theinnermostloop.substack.com/p/welcome-to-february-11-2026
This citation does not imply independent verification of linked-source claims.

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