terminal-bench-sample@2.0
Coding
A sample of tasks from Terminal-Bench 2.0.
Run this task
CLI:
inspect eval inspect_harbor/terminal_bench_sample_2_0 --model openai/gpt-5Python:
from inspect_ai import eval
from inspect_harbor import terminal_bench_sample_2_0
eval(terminal_bench_sample_2_0(), model="openai/gpt-5")Dataset information
| Harbor registry | terminal-bench-sample@2.0 |
| Inspect task | terminal_bench_sample_2_0 |
| Version | 2.0 |
| Samples | 10 |
| Paper | arxiv |
See Task Parameters for the parameter set shared across all Harbor tasks.