Analyzing constrained LLM through PDFA-learning

Alejo Garat; Franz Mayr; Johny Kidd; Juan Pedro da Silva; Mart\'in Iturbide; Mat\'ias Carrasco; Sergio Yovine

Not yet reviewed by Pith; the record is open.

Re-run · record.json Download PDF Read on arXiv ↗

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2406.08269 v2 pith:7AWPWBHM submitted 2024-06-12 cs.FL cs.AIcs.LG

Analyzing constrained LLM through PDFA-learning

Mat\'ias Carrasco , Franz Mayr , Sergio Yovine , Johny Kidd , Mart\'in Iturbide , Juan Pedro da Silva , Alejo Garat This is my paper

classification cs.FL cs.AIcs.LG

keywords analyzingcongruenceconstrainedalgorithmarisecasecopesdefine

verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

0 comments

read the original abstract

We define a congruence that copes with null next-symbol probabilities that arise when the output of a language model is constrained by some means during text generation. We develop an algorithm for efficiently learning the quotient with respect to this congruence and evaluate it on case studies for analyzing statistical properties of LLM.

discussion (0)

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

Minimality of Random Moore Automata under Prefix-Dependent Congruences
cs.FL 2026-06 unverdicted novelty 6.0

Under independent uniform random transitions and labels with agreement probability strictly less than one and at least three admissible symbols each, the induced prefix-dependent congruence on random Moore automata is...