featurebench/featurebench-lite

Coding

Lightweight subset of FeatureBench for cheaper evaluation while preserving model rankings.

← Back to Registry

Run this task

CLI:

inspect eval inspect_harbor/featurebench_lite --model openai/gpt-5

Python:

from inspect_ai import eval
from inspect_harbor import featurebench_lite

eval(featurebench_lite(), model="openai/gpt-5")

Dataset information

Harbor registry featurebench/featurebench-lite
Inspect task featurebench_lite
Latest digest sha256:8cc1a98d3087e46dccce1807f6032dbc12ad054aa8a742e668e31ebc05bb1b4d
Samples 30
Paper arxiv

See Task Parameters for the parameter set shared across all Harbor tasks.