Link the arXiv paper (2609.03350)

Re-chunk test_tail parquet (20K-row groups, page index) for the dataset viewer

Re-chunk test_iid parquet (20K-row groups, page index) for the dataset viewer

Re-chunk val parquet (20K-row groups, page index) for the dataset viewer

Re-chunk train parquet (20K-row groups, page index) for the dataset viewer

Sync with paper v2: subset manifest, comparable corpus sizes

Corpus size in Gemma tokens with SP-token note

Round training-document count

Card: source field name

Card: field names match parquet schema

Author order update

Full author list + team section

Update card links to COPA-AI

ArmWeb manifest

ArmWeb split test_tail

ArmWeb split test_iid

ArmWeb split val

ArmWeb split train

Add release card

initial commit