Where GPQA remains difficult

Aggregate question difficulty across 268 accepted configurations,
without publishing prompts, answer keys, or a model-by-question matrix.

Results 0.1.14Data through Oct 8, 2026

189 retained questions · 3 domains · 9 reportable subcategories · Aggregate-first publication boundary
Method notes

These controls preview a future, separately reviewed question-analysis release. This release intentionally publishes one aggregate cohort only.

Displayed statistics require ≥5 questions and ≥5 configurations.

1. Difficulty by domain

  1. Biology1771.1%71.1%
  2. Chemistry8863.8%63.8%
  3. Physics8480.7%80.7%

Each configuration is weighted equally. Repeated attempts reduce to one per-question mean before aggregation.

2. Subcategories with enough questions to report

4 subcategories suppressed: fewer than 5 questions
SubcategoryQuestionsMean correctnessValue
Astrophysics1374.0%74.0%
Chemistry (general)1972.1%72.1%
Electromagnetism and Photonics674.1%74.1%
High-energy particle physics1484.3%84.3%
Molecular Biology1471.6%71.6%
Organic Chemistry6861.4%61.4%
Physics (general)1879.9%79.9%
Quantum Mechanics2583.4%83.4%
Relativistic Mechanics684.7%84.7%

3. Consistently difficult patterns

1

Organic Chemistry is the hardest reportable subcategory.

61.4% mean correctness.

2

Difficulty remains broad.

The release does not expose Record IDs or per-model correctness.

3

Descriptive, not causal.

These aggregates do not identify why a configuration failed a question.

4. Question disclosure boundary

Included in this release

Domain/subcategory aggregates, counts, definitions, cohort scope, and suppression markers.

Held back pending dedicated review

Record IDs, per-model correctness, prompt text, choices, keys, completions, and reasoning.

5. How to read this page

  1. Attempts per configuration
  2. Per-question configuration mean
  3. Aggregate configurations equally
  4. Suppress small cells before rendering
Download aggregate JSON

Explore the benchmark

Open leaderboard See cleaning impact Read methodology