swtbench-verified@1.0

Coding

SWTBench Verified - Software Testing Benchmark for code generation.

← Back to Registry

Run this task

CLI:

inspect eval inspect_harbor/swtbench_verified_1_0 --model openai/gpt-5

Python:

from inspect_ai import eval
from inspect_harbor import swtbench_verified_1_0

eval(swtbench_verified_1_0(), model="openai/gpt-5")

Dataset information

Harbor registry swtbench-verified@1.0
Inspect task swtbench_verified_1_0
Version 1.0
Samples 433
Paper arxiv

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