Datasetpublic

ColomboAI/C3R-DecisionMix-v1-preview

C3R DecisionMix v1 Preview This is a deterministic synthetic schema preview, not the empirical DecisionMix v1 training corpus and not a benchmark result. It exists so researchers can inspect, load, and validate the record contract before governed traces are collected and released. Each record includes the compiled state, typed questions, candidate and counterfactual actions, baseline action, verifier and task outcomes, latency, cost, risk class, provider identity, and source… See the full description on the dataset page: https://huggingface.co/datasets/ColomboAI/C3R-DecisionMix-v1-preview.

sourceHugging Faceapache-2.0updated 2d agoView on Hugging Face

0likes36downloads

Dataset Card

C3R DecisionMix v1 Preview

This is a deterministic synthetic schema preview, not the empirical DecisionMix v1 training corpus and not a benchmark result. It exists so researchers can inspect, load, and validate the record contract before governed traces are collected and released.

Each record includes the compiled state, typed questions, candidate and counterfactual actions, baseline action, verifier and task outcomes, latency, cost, risk class, provider identity, and source provenance. Splits are immutable functions of record ID and the seed in manifest.json.

Intended use

—validate ingestion and training pipelines;
—prototype typed-decision and calibration tooling;
—review the data contract and contribute fixtures.

Do not use these synthetic labels to claim real-world task accuracy, calibration, latency, cost, or safety performance. The empirical release remains gated on source licensing, held-out split integrity, sufficient calibration support, and publication of raw evaluation evidence.

Rebuild

bash

python scripts/build_decisionmix_preview.py

All preview content is licensed under Apache-2.0. See the repository LICENSE and NOTICE.