Skip to content

Dataset · MatrAIx2026

One million personas with attributes

MatrAIx2026/MatrAIx_Persona_1M

999,847 personas with 1,290 categorical attributes for character generation and analysis.

What's inside

Contains 999,847 persona records, each with 1,290 categorical attributes; 599,847 records are sourced from real data, while 400,000 are synthetic. The data is stored in 10 parquet shards totaling 4.17 GB, with attributes packed using 4-bit codes. Reading requires pyarrow and the persona_codes.schema.json schema, as the datasets library does not open these files. Metadata lists the size as n<1K, which conflicts with the card.

What you can do with it

  • Train text generation based on a persona profile
  • Sample synthetic user profiles
  • Verify decoding of packed tabular attributes
Tasks
Text generation
Size
up to 1K rows
Downloads this month
1.3K
Likes
90

How to load it

from datasets import load_dataset

ds = load_dataset("MatrAIx2026/MatrAIx_Persona_1M")