Hugging Face's logo

Datasets:
OPENGCM
/
DeepArithmark-1

Tasks:
Question Answering
Modalities:
Text
Formats:
json
Languages:
English
Size:
1K - 10K
Tags:
arithmetic
multiple-choice
Synthetic
language-model-evaluation
long-horizon
Libraries:
Datasets
pandas
Polars
License:
Dataset card Data Studio Files
xet
 Community
1
DeepArithmark-1 / README.md

Commit History

4 days ago

  1. Update README.md

    TeamNull
    Verified signature
    b1e2378

  2. Rewrap new score paragraphs; point baselines at the 17-model file

    TeamNull
    Verified signature
    d1cdf1c

  3. Score 1.6-3.1B models: SmolLM3-3B, Qwen3.5-2B, LFM3-1.7B, granite-1B

    TeamNull
    Verified signature
    63de4eb

6 days ago

  1. Update README.md

    TeamNull
    Verified signature
    0ca1246

  2. Restructure card: quick start and scores first

    TeamNull
    Verified signature
    e47dec8

  3. Update README.md

    TeamNull
    Verified signature
    745ec66

  4. Clarify significance claim in baselines

    TeamNull
    Verified signature
    3d84138

  5. Add 2026 sub-1B baseline sweep

    TeamNull
    Verified signature
    5e27d43

  6. Update README.md

    TeamNull
    Verified signature
    7eefd22

  7. Add Deep Arithmark 1: 1,000 long-horizon arithmetic items, generator, scorer, design notes

    TeamNull
    Verified signature
    d765c14