# Source review and mixture rationale

This is the source-selection record for the local v1 build. It is deliberately separate from the normalized data so the model artifact can be released without publishing the derived training set.

## Included

| Source | Role in v1 | Important caveat |
| --- | --- | --- |
| [open-r1/OpenR1-Math-220k](https://huggingface.co/datasets/open-r1/OpenR1-Math-220k) | Main math source; multiple reasoning traces and Math-Verify correctness metadata | The builder uses only a bounded streamed subset and creates evidence views; it does not claim every trace is human-verified. |
| [aladinDJ/ultramix-DPO-annotated](https://huggingface.co/datasets/aladinDJ/ultramix-DPO-annotated) | Balanced general, reasoning, math, coding and instruction-following preferences with curation metadata | Preference pairs are kept as pairwise supervision, not treated as absolute truth. |
| [nvidia/HelpSteer3](https://huggingface.co/datasets/nvidia/HelpSteer3) | Human-annotated multi-turn preference coverage across general, STEM, code and multilingual domains | Overall preference has a three-point direction scale; ties are skipped. |
| [coseal/CodeUltraFeedback_binarized](https://huggingface.co/datasets/coseal/CodeUltraFeedback_binarized) | Code-specific preference and rating signal | Identical chosen/rejected pairs are skipped. |
| [Skywork/Skywork-Reward-Preference-80K-v0.2](https://huggingface.co/datasets/Skywork/Skywork-Reward-Preference-80K-v0.2) | Additional general reward-preference diversity | The source has limited task metadata, so rows use the generic rubric and preserve the upstream source tag. |
| [R-I-S-E/RISE-Judge-SFT-20K](https://huggingface.co/datasets/R-I-S-E/RISE-Judge-SFT-20K) | Judge-style comparisons with a generated rationale and position-swap verification | The builder extracts the user question and A/B candidates, discarding the long judge rationale from the scored candidate. |
| [iknow-lab/JudgeBias-DPO-RefFree](https://huggingface.co/datasets/iknow-lab/JudgeBias-DPO-RefFree) | Bounded anti-shortcut/materials-science pairs for representation invariance and injected-error detection | The builder uses only the stable `validation_1000.parquet` file because the full upstream parquet set currently has inconsistent schemas. |
| [Anthropic/hh-rlhf](https://huggingface.co/datasets/Anthropic/hh-rlhf) | Multi-turn helpfulness and harmlessness preferences from human feedback | The default Hub config exposes transcript strings, so the builder extracts the shared dialogue prefix and final assistant response. The source contains sensitive content; keep the derived data private unless separately cleared. |
| [stanfordnlp/SHP](https://huggingface.co/datasets/stanfordnlp/SHP) | Naturally occurring human-written preferences across 18 question/instruction domains | Labels are collective Reddit preferences, not factual correctness. The builder filters weak score-ratio pairs and records subreddit metadata. The card discusses Reddit API terms rather than granting a simple redistribution license. |
| [AIPlans/PKU-SafeRLHF-RLHF](https://huggingface.co/datasets/AIPlans/PKU-SafeRLHF-RLHF) | Safety-aware preference pairs with a derived margin for confidence weighting | CC-BY-NC-4.0 is inherited from PKU SafeRLHF; this is suitable for the private research build, not automatically for commercial redistribution. |
| [SUSTech-NLP/MixReward](https://huggingface.co/datasets/SUSTech-NLP/MixReward) | Multilingual preference coverage across 103 languages and six broad domains | Apache-2.0 card, but the mixture aggregates multiple upstream sources; retain the language and source-subset metadata when auditing or redistributing. |
| [project-themis/Themis-CodePreference](https://huggingface.co/datasets/project-themis/Themis-CodePreference) | Multi-criterion code preference signal spanning correctness, efficiency, security and maintainability across eight languages | Apache-2.0 card. Aspect and language IDs are mapped into the row instruction and metadata; the source itself is already a curated mixture, so the quota stays bounded to avoid code dominance. |

## Explored but not in the default build

| Source | Decision |
| --- | --- |
| [OpenAI PRM800K](https://github.com/openai/prm800k) | Deliberately not used as the main math source because v1 is intended to be richer than simple step labels. |
| Math-Shepherd | Useful process-supervision reference, but not included in the default mixture; its automated step labels would add another label regime before the core schema is stable. |
| Full JudgeBias train/validation parquet set | Not used because the Hub parquet files currently expose inconsistent columns and fail the streaming cast check; only the stable 999-row `validation_1000.parquet` slice is included. |
| [amazon/CodePrefBench](https://huggingface.co/datasets/amazon/CodePrefBench) | Kept as an evaluation candidate rather than training data; its Hub card identifies a non-commercial license and an evaluation/test orientation. |
| [cracklinoatbran/reward_hacking_policy_1073](https://huggingface.co/datasets/cracklinoatbran/reward_hacking_policy_1073) | Prompt-only reward-hacking evaluation set. It needs fresh candidate generations and verified hack/legitimate labels, so it is reserved for a later anti-shortcut generation stage. |
| [hlyn-labs/prompt-injection-judge-dataset-v1](https://huggingface.co/datasets/hlyn-labs/prompt-injection-judge-dataset-v1) | Security-judge conversations are not directly compatible with the v1 candidate-level correctness contract; reserve for a later adversarial/instruction-injection slice. |

## Mixture properties of the generated artifact

The expanded build contains 53,955 unique candidate rows in three group-isolated Parquet splits:

- 16,000 absolute math-verification rows.
- 37,955 pairwise preference/judge rows across general chat, multilingual, safety, STEM, code, materials science, and math/judgment sources.
- 41,959 rows with no evidence, 3,976 with a reference solution, 4,014 with privileged context, and 4,006 with both.
- 48,607 train, 2,697 validation, and 2,651 test rows.
- 53,955 unique `example_id` values and 27,219 unique `group_id` values.
- 12 source datasets, with 160,198,172 bytes across the three Parquet shards.

The source datasets are not all licensed identically. This folder is therefore a private local artifact by default; check each upstream dataset's current license and terms before redistributing derived data. The model-weight release should document the source mixture without publishing these Parquet files unless redistribution is separately cleared.