0

Agency Bench RL Env (Prime Intellect)

Fresh

HumanAgencyBench: Benchmark measuring AI assistants' support for human agency across 6 dimensions (3000 prompts, LLM-as-judge)

Type
RL Env
Publisher
Prime Intellect
Tags
Human Agency
Runtime
single-turn
License
unknown
Size
v0.1.0
Published
Nov 2025
Canonical
app.primeintellect.ai/dashboard/environments/primeintellect/agency-bench

Cite

Notes

Only stored in your browser.

Hub: primeintellect/agency-bench · v0.1.0 · team Task: single-turn · Parser: No Tags: benchmark llm-as-judge human-agency eval

Install

prime env install primeintellect/agency-bench
# or via pip:
uv pip install agency-bench --extra-index-url https://hub.primeintellect.ai/primeintellect/simple/

Dependencies

verifiers==0.1.5, openai>=1.0.0, datasets

Python: >=3.10

Provenance