biopharma-bench/biopharma-bench

Ten assignments at Havenor Therapeutics, a fictional company that makes sterile injectable drugs. The company date is 27 September 2026. An agent takes an employee’s seat and writes the document that seat is responsible for.

← Back to Registry

Run this task

CLI:

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

Python:

from inspect_ai import eval
from inspect_harbor import biopharma_bench

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

Dataset information

Harbor registry	biopharma-bench/biopharma-bench
Inspect task	biopharma_bench
Latest digest	sha256:8975dab9154ee4fdc053650f6ae498ade4e80fcd1f669c4ab2d048bb990faf7c
Samples	10
Source	https://huggingface.co/datasets/raycasterai/biopharma-bench

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