datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
cwv-strategy-mining-dataset
CWV Strategy Mining Dataset
Real, merged GitHub PRs mined for measured Core Web Vitals (CWV)
techniques, plus every intermediate pipeline artifact through final
generated playbook candidates. Produced by
cwv-playbook-miner,
scanning the full public GH Archive event stream. Every decision in the
pipeline runs on real PR text (title, body, diff, comments, reviews) — never
a short phrase, never a bare similarity threshold. See the repo's
SESSION_NOTES.md for the full chronological… See the full description on the dataset page: https://huggingface.co/datasets/Ayush-Singh/cwv-strategy-mining-dataset.latent-mining
Latent Mining
Latent Mining is a benchmark-construction method for scientific-agent tasks where useful public evidence diverges from a withheld verifier-backed outcome. The resulting tasks test whether agents can make calibrated scientific triage decisions under incomplete information.
This dataset contains the first public biology subset: 165 cross-locus regulatory-edit triage tasks. Each task asks an agent to choose among candidate noncoding edits for a specified assay and… See the full description on the dataset page: https://huggingface.co/datasets/Jarrodbarnes/latent-mining.albedo-mining-scripts
albedo-mining-scripts
Training scripts and shared-fail SFT/DPO mix for Albedo SN97 (Qwen3.6-35B-A3B),
rebuilt for king 119. See RUNBOOK.md for the GPU run.
Start weights:
comguys/albedo-qwen3.6-35b-rzzi5i6@0c5dd11b7be00aae874e63d53d0e454f81c3e4a1
datasets/mix/sft.jsonl — 1679 upsampled SFT rows (reference trajectories)
datasets/mix/dpo.jsonl — 332 unique DPO pairs (chat vs chat; rejected is king 119 when the eval has those turns)
evals/ — source eval artifacts, including 119's… See the full description on the dataset page: https://huggingface.co/datasets/athena2634/albedo-mining-scripts.orbital-mining-corpus-seeded
Orbital Mining Corpus — Seeded
What's inside — The 848 Orbital Mining Corporation technical documents plus 16 new synthetic OMC docs (incident reports, ops logs, handovers, test reports, maintenance procedures).
Where it comes from — orbital-mining-corpus, extended by CROSS-SEEDING: teacher-written docs (Scaleway, gpt-oss-120b) anchored to real telemetry values from the Mars facts table.
How it was built / modified — Strict grounding — only candidates whose telemetric values… See the full description on the dataset page: https://huggingface.co/datasets/atenareply/orbital-mining-corpus-seeded.