{
 "id": "Infektyd/syntra-evals-dataset",
 "owner": "Infektyd",
 "name": "syntra-evals-dataset",
 "downloads": 55,
 "likes": 0,
 "last_modified": "2026-04-02T11:57:41.000Z",
 "stated_licence": "mit",
 "stated_from": "tag",
 "licence_name": null,
 "licence_link": null,
 "file_licence": null,
 "licence_files": [],
 "sources": [],
 "gated": false,
 "tasks": [
  "text-generation",
  "text-classification"
 ],
 "citation": "The `cmt-physics` split draws on problems from the CMT-Benchmark ([arXiv 2510.05228](https://arxiv.org/abs/2510.05228), [JVRoggeveen/cmt_benchmark](https://huggingface.co/datasets/JVRoggeveen/cmt_benchmark), MIT licensed). Gold standard answers were available in the paper; the evaluation tooling and grading framework were built from scratch alongside the testing.",
 "signal": "permissive",
 "reason": "The declared licence (mit) is classified as permissive.",
 "revision": "c3a32dc144a62fd64cb18397252c153a2a7d9d5e",
 "stated_evidence": "https://huggingface.co/datasets/Infektyd/syntra-evals-dataset/blob/c3a32dc144a62fd64cb18397252c153a2a7d9d5e/README.md",
 "file_evidence": "https://huggingface.co/datasets/Infektyd/syntra-evals-dataset/tree/c3a32dc144a62fd64cb18397252c153a2a7d9d5e",
 "card_nc": false
}