alea-institute/alea-defined-terms

ALEA defined terms

Defined-term tasks over contract text: list the defined terms, quote a term's definition exactly, say whether a term is defined (or absent, or used without a definition), and list terms used but never defined.

16,408 documents, 5,609,092 tokens (alea-v2b-sw32k, the ALEA R4 tokenizer, counting each document's start and end tokens).

Part of the ALEA R4 training mix

This dataset contains exactly the documents of this family that ALEA's R4 model trained on in its stable pretraining mixture (second data stage, one pass each). Every row is byte-identical to the trained text; text_sha256 is the SHA-256 of text.

Held-out evaluation items generated by the same code are not part of this release.

Composition

task	documents	tokens
define	4,174	1,375,774
is_defined	4,138	1,265,736
undefined	4,137	1,636,150
list_terms	3,959	1,331,432

5,643 documents use ALEA's chat template.

How it was made

Written by deterministic code; no language model was used. Real excerpts are windows of 4–10 consecutive definition blocks from material-contract exhibits (EX-10) in SEC EDGAR filings, kept only when the quoted terms are exactly the defined terms. Synthetic agreements are built from typed slots with a closed term bank and planted undefined distractors.

Sources

• SEC EDGAR EX-10 exhibits (public record)
• Invented agreements from typed templates

Verification

• Synthetic agreements assert that every mid-sentence capitalised word is a defined term or a planted undefined term, and that definitions are closed transitively.
• Real excerpts are complete definition blocks ending in '.' or ';'.
• Decontamination: training rows that shared a word 13-gram with a held-out evaluation item of this or any sibling family were removed, and rows overlapping ALEA's evaluation sets were removed or logged according to ALEA's decontamination policy.
• Sampled records were reviewed blind by a separate reviewer who recomputed or re-derived the answers; defect patterns found in review were fixed in the generator, or the affected rows dropped, before the final build.

Columns

column	content
id	row id from the generator (not always unique: a few families reuse an id across variants, and a handful of exact duplicate documents were trained twice and appear twice)
text	the document exactly as trained; chat documents use ALEA's chat template, whose control tokens appear as text (`<
task	task kind
phase	stable (trained in R4's stable pretraining mixture)
generator	what wrote the row (deterministic = code only)
group_id	source passage / record group; all variants of one group share it
prompt_chars	length in characters of the request part of text (0 for plain documents; null where not recorded)
n_tokens	tokens under alea-v2b-sw32k, including the document start and end tokens
text_sha256	SHA-256 of text
meta	JSON: the generator's structured fields (source, presentation, answer form, gold values, checks)

Known limitations

• The documents are generated and more repetitive than natural text.

License

CC-BY-4.0 for this compilation and the generated text. Public-domain and public-record source text keeps its status; Wikidata is CC0; OpenGloss and SHELF are CC-BY-4.0.

Published by the ALEA Institute.