English

An $\mathbf{L^*}$ Algorithm for Deterministic Weighted Regular Languages

Computation and Language 2024-12-20 v2

Abstract

Extracting finite state automata (FSAs) from black-box models offers a powerful approach to gaining interpretable insights into complex model behaviors. To support this pursuit, we present a weighted variant of Angluin's (1987) L\mathbf{L^*} algorithm for learning FSAs. We stay faithful to the original algorithm, devising a way to exactly learn deterministic weighted FSAs whose weights support division. Furthermore, we formulate the learning process in a manner that highlights the connection with FSA minimization, showing how L\mathbf{L^*} directly learns a minimal automaton for the target language.

Keywords

Cite

@article{arxiv.2411.06228,
  title  = {An $\mathbf{L^*}$ Algorithm for Deterministic Weighted Regular Languages},
  author = {Clemente Pasti and Talu Karagöz and Anej Svete and Franz Nowak and Reda Boumasmoud and Ryan Cotterell},
  journal= {arXiv preprint arXiv:2411.06228},
  year   = {2024}
}