Metraipolis › archive › dc_6312d3fa9bee

How to get the best and fairest performance out of your ASR system using activation steering — arXiv:2610.26994

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.26994
License (as archived)
CC0-1.0
Archived
2026-10-21T07:08:46.596Z
SHA-256 of the archived text
78e216d893a7b5a1e0b7db1090c239ca38a566b185972d176bd6b315d2d3d179
Stable URL of this page
https://metraipolis.com/documento/dc_6312d3fa9bee

How to cite

How to get the best and fairest performance out of your ASR system using activation steering — arXiv:2610.26994 — arXiv (en), https://arxiv.org/abs/2610.26994 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_6312d3fa9bee (Metraipolis archive dc_6312d3fa9bee, sha256 78e216d893a7b5a1e0b7db1090c239ca38a566b185972d176bd6b315d2d3d179)
BibTeX
@misc{dc6312d3fa9bee,
  title = {How to get the best and fairest performance out of your ASR system using activation steering — arXiv:2610.26994},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.26994},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_6312d3fa9bee; sha256 78e216d893a7b5a1e0b7db1090c239ca38a566b185972d176bd6b315d2d3d179},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.26994. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Activation steering for ASR systems is an underexplored technique in improving both ASR performance and demographic fairness. We investigate two forms of activation steering: paired activation subtraction, and supervised optimization. We find the former to be ineffective at ameliorating ASR performance whatsoever, while the latter is quite effective at deriving steering vectors (SVs) that improve at once overall ASR performance and demographic fairness along such diverse axes as gender, age, and dialect. We also explore what types of data---from the most to the least personal---are necessary for calculating or training SVs. First, we show significant performance in performance from training SVs using an eclectic group of speakers from the same dataset. Furthermore, we find that vectors trained using acoustically similar speakers leads to ASR performances that are competitive or superior to vectors trained on the same speaker, or speakers from the same labeled demographic group. This finding shows that we can harness activation steering for few-shot adaptation given access to a speaker's demography, or zero-shot adaptation to unseen speakers without the need for any labels at all.

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/2610.26994 — 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 2026-10-21T07:05:58.094Z dalla pagina https://arxiv.org/list/cs.CL/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).