scale-ai/swe-atlas-tw

Coding

SWE-Atlas - Test Writing – A benchmark of comprehensive test writing problems for coding agents. Checkout for instructions on running it.

← Back to Registry

Run this task

CLI:

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

Python:

from inspect_ai import eval
from inspect_harbor import scale_ai_swe_atlas_tw

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

Dataset information

Harbor registry scale-ai/swe-atlas-tw
Inspect task scale_ai_swe_atlas_tw
Latest digest sha256:1fe41edad7c1cc96925f100be0d314804cc45f55a9a5b7c8583354f4bca30b44
Samples 90
Source https://github.com/scaleapi/SWE-Atlas

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