Metraipolis › archive › dc_ca3c8cb07dd7

When Can You Prune Your Network? A Study of Intermediate Neurons in Multilingual Speech Parsing — arXiv:2610.11520

Source: arXiv (en)
Original: https://arxiv.org/abs/2610.11520
License (as archived): CC0-1.0
Archived: 2026-10-09T06:06:29.767Z
SHA-256 of the archived text: dd15d07e946efc5faf9d728915506592dd70c0adbddb08c167a0d2c632b119d6
Stable URL of this page: https://metraipolis.com/documento/dc_ca3c8cb07dd7

How to cite

When Can You Prune Your Network? A Study of Intermediate Neurons in Multilingual Speech Parsing — arXiv:2610.11520 — arXiv (en), https://arxiv.org/abs/2610.11520 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_ca3c8cb07dd7 (Metraipolis archive dc_ca3c8cb07dd7, sha256 dd15d07e946efc5faf9d728915506592dd70c0adbddb08c167a0d2c632b119d6)

BibTeX

@misc{dcca3c8cb07dd7,
  title = {When Can You Prune Your Network? A Study of Intermediate Neurons in Multilingual Speech Parsing — arXiv:2610.11520},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.11520},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_ca3c8cb07dd7; sha256 dd15d07e946efc5faf9d728915506592dd70c0adbddb08c167a0d2c632b119d6},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.11520. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

End-to-end speech parsing, a task recently proposed, consists in predicting both the transcription and the syntactic tree for a spoken utterance. Existing architectures for speech parsing often utilise intermediate neural networks. In this work, we examine the effectiveness of intermediate neural networks (NN) for parsing, and, specifically, what role do they play. We introduce a simpler end-to-end architecture for speech parsing, where we remove these intermediate NN units, reducing the parameters by 12%, while achieving comparable or better performance than prior method on both automatic speech recognition (ASR) and parsing. We demonstrate that intermediate NN units help reduce the representational gap when the pre-trained encoder is frozen. We do a comprehensive evaluation of speech parsing on French, and medium-low resource languages Slovenian and Naija. We further investigate the impact of the training data size and intermediate layers of the pretrained speech encoder on speech parsing.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.CL, nuovi lavori del giorno).
Categorie arXiv: Computation and Language (cs.CL).
Pagina del lavoro: https://arxiv.org/abs/2610.11520 — 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 2026-10-09T05:47:15.212Z dalla pagina https://arxiv.org/list/cs.CL/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).