alea-institute/alea-summary-control

ALEA controlled summaries

Summaries of record documents under explicit controls: one sentence, N sentences, N bullets, N numbered takeaways, a headline plus one sentence, a TL;DR line, or one paragraph under a word cap, in a formal or technical-legal register. Each row is an instruction, the source document and the summary.

44,688 documents, 22,869,710 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. Chat-format rows that the generator reserved for supervised fine-tuning are not included.

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

Composition

task	documents	tokens
sumctl:one_sentence:formal	5,707	2,686,999
sumctl:one_sentence:technical_legal	5,574	2,627,434
sumctl:tldr:technical_legal	5,549	2,702,652
sumctl:tldr:formal	5,331	2,588,077
sumctl:n_sentences:formal	4,059	2,175,242
sumctl:n_sentences:technical_legal	3,845	2,081,504
sumctl:n_numbered:technical_legal	2,859	1,634,659
sumctl:n_numbered:formal	2,819	1,612,527
sumctl:n_bullets:technical_legal	2,047	1,188,772
sumctl:n_bullets:formal	1,829	1,059,066
sumctl:paragraph_cap:formal	1,494	728,885
sumctl:paragraph_cap:technical_legal	1,487	724,840
sumctl:headline_sentence:formal	1,375	691,856
sumctl:headline_sentence:technical_legal	713	367,197

How it was made

Qwen3.8-27B (open weights, served locally) wrote the summaries of documents from the record-documents family; the request never gives it the values to include.

Sources

• ALEA record documents (invented records)

Verification

• Fail-closed checks: exact counts (sentences by nupunkt, bullets and items by line structure, words) and caps; must-keep facts with exact numbers and dates; conditions, exceptions and amendments kept; no out-of-source legal or payment terms; payer order; status and negation polarity; a copy cap on verbatim spans.
• 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	the generator's placement tag: pretraining for plain documents, midtraining for instruction-framed tasks; every row here was 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

• Only formal and technical-legal registers are included.
• 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.