Datasets:

WithinUsAI
/

All_Known_Species_Index_25k

language: en
license: cc-by-4.0
task_categories:
  - text-generation
tags:
  - taxonomy
  - species
  - ncbi
  - biodiversity
  - withinusai
size_categories:
  - 10K<n<100K
pretty_name: WithinUsAI All Known Species Index 25k (NCBI Taxonomy Sample)
dataset_info:
  features:
    - name: prompt
      dtype: dict
    - name: response
      dtype: string
    - name: meta
      dtype: dict
    - name: fact_sources
      dtype: list
    - name: artifacts
      dtype: dict

WithinUsAI/All_Known_Species_Index_25k — Master Scholars Academics (25k)

A real-species index dataset sampled from NCBI Taxonomy (taxdump) at the species rank, formatted for Tiny-Recursive-Model-friendly fine-tuning.

What this is (and is not)

This dataset contains 25,000 real species records sampled from the NCBI taxonomy graph (rank=species).
It does not enumerate every described species name (the full list is far larger and updated continuously).
If you want “all known species” as a full index, generate it from the same official NCBI taxdump or other backbones (Catalogue of Life, GBIF, WoRMS). This repo provides the schema and can be scaled.

Task

Each row asks the model to return a taxonomy classification JSON for the given taxid + scientific name, with keys among:

superkingdom, kingdom, phylum, class, order, family, genus, species

Output contract: JSON only.

Source

Official data backbone:

NCBI Taxonomy FTP: taxdump.tar.gz
https://ftp.ncbi.nlm.nih.gov/pub/taxonomy/taxdump.tar.gz

Accessed: 2026-01-04T03:02:03Z

Schema

prompt.messages[]: system + user
response: JSON string containing taxid, scientific_name, and classification
meta: ids, task, ranks, tags
fact_sources: official anchor
artifacts: output contract and token hints

Example

{
  "prompt": {
    "messages": [
      {"role": "system", "content": "You are WithinUsAI Master Scholars Taxonomy Indexer. Output must be valid JSON only, no extra text."},
      {"role": "user", "content": "NCBI Taxonomy species indexing task..."}
    ]
  },
  "response": "{\"taxid\": 9606, \"scientific_name\": \"Homo sapiens\", \"classification\": {\"superkingdom\": \"Eukaryota\", \"kingdom\": \"Metazoa\", \"phylum\": \"Chordata\", \"class\": \"Mammalia\", \"order\": \"Primates\", \"family\": \"Hominidae\", \"genus\": \"Homo\", \"species\": \"Homo sapiens\"}}",
  "meta": {
    "id": "WUA_ALL_SPECIES_25K_000001",
    "domain": "all_known_species_index",
    "subdomain": "ncbi_taxonomy::species",
    "task": "taxonomy_indexing",
    "truth_mode": "verifiable_from_ncbi_taxdump",
    "wrapper": "json_only",
    "created_utc": "2026-01-04T03:02:03Z",
    "license": "cc-by-4.0"
  }
}

Citation

If you use this dataset in research or releases, cite it as: Within Us AI — WithinUsAI/All_Known_Species_Index_25k (NCBI Taxonomy Sample).