Claim · 1996-10 · event · Stated in source · Supported

Hu, Turin and Brown (ICSLP 96) say that large language models are often required to achieve low perplexity with adequate coverage in speech and handwriting recognition.

One recorded claim. The claim panel gives its qualification, the source passages with their place in each source, and how it was checked.

What it connects

Language Modeling with Stochastic Automata → uses the name →

Where it appears

No collection page holds this claim. It is listed with its records in the explorer.

Claim and evidence

Language Modeling with Stochastic Automatauses the name ↓

Machine-checked, not reviewed by an editorStated in sourceSupported

Supporting evidence: primary · 1 source

Date: 1996-10 · event Records dated 1996 →

Scope and limitsEarliest located use by the recorded search; not evidence of coinage. The sense is statistical (stochastic language models distinguished by size), as for Brants and others (2007).

The evidence 01

primary sourceSupports the claim

Language Modeling with Stochastic Automata

ISCA Archive (ICSLP 96 proceedings) · Publication date unknown

Cite the original source

Cite the original document, not Selekoa. Check these details against it.

Source excerpt

“large language models are often required to achieve low model perplexity (or entropy) and still have adequate language coverage.”

Location in source PDF p. 1, abstract

How it was read Retrieved directly from the publisher

Machine-checked · 2026-10-31

Source details and check notes

Authors: Jianying Hu; William Turin; Michael K. Brown
Version: Retrieved for the candidate pack on 2026-10-29 (SHA-256 e37289026474f08587a9ece70be1ea9e6d70939d38958df6930426a665222303). Proceedings paper as hosted by the ISCA Archive (doi 10.21437/ICSLP.1996-79); retrieved by the same-family audit (pass B).
Canonical address: Open the canonical source ↗
Accessed: 2026-10-29
Language: English (en)
Attribution: Jianying Hu, William Turin, Michael K. Brown, “Language Modeling with Stochastic Automata”.
Reuse: No open redistribution licence asserted. Bibliographic metadata, editorial paraphrases and short attributed quotations (at most 50 words each). Recheck terms before wider redistribution.
Check notes: Candidate pack passage, Kinds of AI and technical history (research/technical-history). The quotation (19 words) matched the text retrieved for the candidate pack on 2026-10-29. Matching ignored differences of whitespace, case, quotation marks, dashes, PDF line-break hyphens and Unicode compatibility forms, so the source’s typography can differ. A cross-family machine check (31 October 2026, item 109) re-read the record’s sources and found that they support what the record states. Its point that the conference header dates the conference association, not a presentation was applied at admission. No human editorial review.

Status and meaning

Editorial status. Not reviewed by an editor. An exploratory record that no editor has accepted. This status does not say whether a person has read any linked passage. Each evidence record above shows how it was checked.

Relation. A passage of the document uses a name for a family of methods or a term, in the sense the concept’s naming criterion describes or, where labelled, in another sense. The record documents the use; it is not evidence that the name began there, and it implies no influence on or from another text.

Date. ICSLP 96, Philadelphia, 3–6 October 1996, as the paper’s title-page header states; this is the conference association, not an independently verified day of presentation or publication.

Record ID th-term-llm-hu-1996