gpqa-diamond@1.0
Science
Biology
Chemistry
Physics
Knowledge
GPQA Diamond subset: 198 graduate-level multiple-choice questions in biology, physics, and chemistry for evaluating scientific reasoning.
Run this task
CLI:
inspect eval inspect_harbor/gpqa_diamond_1_0 --model openai/gpt-5Python:
from inspect_ai import eval
from inspect_harbor import gpqa_diamond_1_0
eval(gpqa_diamond_1_0(), model="openai/gpt-5")Dataset information
| Harbor registry | gpqa-diamond@1.0 |
| Inspect task | gpqa_diamond_1_0 |
| Version | 1.0 |
| Samples | 198 |
| Paper | arxiv |
See Task Parameters for the parameter set shared across all Harbor tasks.