scale-ai/swe-atlas-tw@1.0

Coding

SWE-Atlas Test Writing benchmark that evaluates AI agents’ ability to write comprehensive unit tests.

← Back to Registry

Run this task

CLI:

inspect eval inspect_harbor/scale_ai_swe_atlas_tw_1_0 --model openai/gpt-5

Python:

from inspect_ai import eval
from inspect_harbor import scale_ai_swe_atlas_tw_1_0

eval(scale_ai_swe_atlas_tw_1_0(), model="openai/gpt-5")

Dataset information

Harbor registry scale-ai/swe-atlas-tw@1.0
Inspect task scale_ai_swe_atlas_tw_1_0
Version 1.0
Samples 90
Source https://github.com/scaleapi/SWE-Atlas

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