Is this your dataset? Claim it with the Hugging Face account that owns it.

agentic-moral-alignment/
mtma

Name: mtma
Creator: agentic-moral-alignment

other

machiavelli

reinforcement-learning

moral-alignment

environment-analysis

Views

No views yet

mtma — run artifacts

Artifacts from the Multi-Turn Moral Alignment project.

records/structure3/ — MACHIAVELLI env-structure rollouts (v3): 20,123 start points, 2.57M random-policy episodes over a game x depth x walk-seed grid. See records/structure3/README.md for the layout, record schema and how to extract. records/logs/ — collection logs for the above. runs/ — training-run artifacts (per-run logs, metrics, rollout dumps).

records/ is analysis data; runs/ is training output.… See the full description on the dataset page: https://huggingface.co/datasets/agentic-moral-alignment/mtma.