compilebench@1.0
Coding
Version 1.0 of CompileBench, a benchmark on real open-source projects against dependency hell, legacy toolchains, and complex build systems.
Run this task
CLI:
inspect eval inspect_harbor/compilebench_1_0 --model openai/gpt-5Python:
from inspect_ai import eval
from inspect_harbor import compilebench_1_0
eval(compilebench_1_0(), model="openai/gpt-5")Dataset information
| Harbor registry | compilebench@1.0 |
| Inspect task | compilebench_1_0 |
| Version | 1.0 |
| Samples | 15 |
| Paper | arxiv |
See Task Parameters for the parameter set shared across all Harbor tasks.