Oriel — AI & technology worth noticing, every day.

KeepaceAI/MoveBench

Dataset · https://huggingface.co/datasets/KeepaceAI/MoveBench

MoveBench A sports and health coaching benchmark · MoveBench Team Paper · HF Paper Page · Company GitHub · Project / Leaderboard · Submission Guide Dataset summary This Hugging Face distribution contains 80 Chinese evaluation questions only: 50 open-ended questions, 11 single-select questions, and 19 multi-select questions. It covers plan generation (PG, 26), sports data analysis (DA, 29), and professional knowledge question answering (QA, 25). The current… See the full description on the dataset page: https://huggingface.co/datasets/KeepaceAI/MoveBench.

First seen by Oriel on .

AI summary

MoveBench is a gated Chinese benchmark subset containing 80 questions for evaluating language models on sports and health coaching tasks through maintainer-scored submissions.

  • The subset includes 50 open-ended, 11 single-select, and 19 multi-select questions spanning plan generation, sports data analysis, and professional knowledge question answering.
  • Public files exclude rubrics, reference answers, correct options, Judge prompts, API configuration, and scoring code; maintainers score submitted answer files and publish aggregate results.
  • The full paper evaluates 15 models on 809 questions using 5,012 rubric-level criteria, so its scores cannot be reused for this public subset.
  • Access is automatically approved after submitting a Hugging Face form, and the dataset can be loaded with `load_dataset("KeepaceAI/MoveBench", split="test", token=True)`.
  • The CC BY 4.0 dataset has no training split and should be used for evaluation, with disclosure of prior exposure or training involving its materials.

#llm evaluation #sports technology #health coaching #hugging face

Generated by gpt-5.6-sol. Check the original for details.

Metric trends

Last 30 days · Up to 120 recorded samples per source and metric, with reposts kept separate · First and latest retained

Likes

Total likes reported by the source.

  • Hugging Face Datasets: 16, latest observation ; +2 over observed period since

Downloads (30d)

Rolling 30-day downloads. A change in this value is not new downloads during the observed period.

  • Hugging Face Datasets: 764, latest observation ; +178 over observed period since

Aliases

SourceExternal IDURL
Hugging Face DatasetsKeepaceAI/MoveBenchhttps://huggingface.co/datasets/KeepaceAI/MoveBench Read discussion