Metraipolis › archive › dc_bcdfde29887a

Learning of Tree Automata applied to Neural Language Acceptors — arXiv:2611.10147

Source
arXiv (en)
Original
https://arxiv.org/abs/2611.10147
License (as archived)
CC0-1.0
Archived
2026-11-10T18:42:16.134Z
SHA-256 of the archived text
e8654aa811530076c49033f89ad18ed7aa3f62abffef9e04102979a42f2831d0
Stable URL of this page
https://metraipolis.com/documento/dc_bcdfde29887a

How to cite

Learning of Tree Automata applied to Neural Language Acceptors — arXiv:2611.10147 — arXiv (en), https://arxiv.org/abs/2611.10147 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_bcdfde29887a (Metraipolis archive dc_bcdfde29887a, sha256 e8654aa811530076c49033f89ad18ed7aa3f62abffef9e04102979a42f2831d0)
BibTeX
@misc{dcbcdfde29887a,
  title = {Learning of Tree Automata applied to Neural Language Acceptors — arXiv:2611.10147},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.10147},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_bcdfde29887a; sha256 e8654aa811530076c49033f89ad18ed7aa3f62abffef9e04102979a42f2831d0},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.10147. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

We extract visibly pushdown grammars (VPGs) from neural language models in a fully black-box setting. Our work builds on the VPL* framework, which learns VPGs from recurrent networks by exploiting access to the target's internal state. First, we replace the white-box equivalence oracle with PAC sampling over trees. The resulting algorithm applies unchanged to transformer language models and, more generally, to any acceptor exposing only binary decisions. Second, we further observe that the original framework only ever queries the target on well-formed sequences that can be accepted by the most permissive VPG (called BParse) with the target's ranked alphabet. That is, VPL* is blind to the target's behavior outside BParse. Thus, we propose to sample trees directly instead of sequences and call the target with all the resulting sequences in order to provide an error estimate of how far the target deviates from being a VPG. Last but not least, we develop a learner that captures a tree automaton describing discovered behaviors of the target outside BParse, which results in a larger learnable space. We evaluate the approach on several cases, including transformers trained with Dyck grammars and synthetic targets designed to be outside BParse.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno).
Categorie arXiv: Computation and Language (cs.CL) ; Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2611.10147 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-11-10T18:37:37.575Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).