Curated datasets

Policy document chunks (retrieval units with citations)

vlada_curated.fact_policy_doc_chunk·live·curated
Rows
N/A
Columns
8
History
N/A
Card
nocolumns only
Typed tools
0

What it is

  1. 1Chunked policy text for retrieval: doc_id, chunk_no, anchor, citation, text.
  2. 2The retrieval grain for document-grounded policy answers (every claim ties to a cited chunk).

Grain

KeyGrain
N/A
  1. 1One row per (doc_id, chunk_no) — a citation-anchored text chunk.

Columns (8)

ColumnTypeMeaning
doc_idstring
statestring
chunk_noint
anchorstring
citationstring
source_urlstring
vintagestring
textstring

Query it

describe_data_source(source_fqn="vlada_curated.fact_policy_doc_chunk")     # the card, live
query_substrate(sql="SELECT * FROM vlada_curated.fact_policy_doc_chunk WHERE <filter> LIMIT 100")
ItemValue
Typed tools that read itnone; raw SQL
Identity keydoc_id
Joinhow_to_join(source_fqn="vlada_curated.fact_policy_doc_chunk")
Raw doorAthena (read-only via query_substrate): vlada_curated.fact_policy_doc_chunk

This page as markdown: /docs-md/data/vlada_curated_fact_policy_doc_chunk.md · every page in one file: /llms-full.txt