Metraipolis › archive › dc_3198c05ba34f

Reducing Hallucinated Transcripts in Whisper via Hallucination Space Projection — arXiv:2609.04561

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.04561
License (as archived)
CC0-1.0
Archived
2026-10-16T22:20:46.251Z
SHA-256 of the archived text
d7d382589e7ed1d798978bad304bf51ba84ede53fbd41edd1c6baa284419e694
Stable URL of this page
https://metraipolis.com/documento/dc_3198c05ba34f

How to cite

Reducing Hallucinated Transcripts in Whisper via Hallucination Space Projection — arXiv:2609.04561 — arXiv (en), https://arxiv.org/abs/2609.04561 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_3198c05ba34f (Metraipolis archive dc_3198c05ba34f, sha256 d7d382589e7ed1d798978bad304bf51ba84ede53fbd41edd1c6baa284419e694)
BibTeX
@misc{dc3198c05ba34f,
  title = {Reducing Hallucinated Transcripts in Whisper via Hallucination Space Projection — arXiv:2609.04561},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.04561},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_3198c05ba34f; sha256 d7d382589e7ed1d798978bad304bf51ba84ede53fbd41edd1c6baa284419e694},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.04561. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Whisper is a widely used foundation model for automatic speech recognition (ASR), but its generative decoder can produce fluent hallucinated transcripts for inputs containing little or no speech. We propose a training-free, inference-time method to reduce these hallucinations using low-rank projection of decoder activations. A compact hallucination-associated subspace is estimated from non-speech calibration data, and decoder hidden states are projected away from this subspace during inference. We evaluate two variants: always-on, which applies projection to all inputs, and gated, which applies it only when Whisper predicts that an input is likely non-speech. Across non-speech benchmarks, always-on projection reduces average hallucination rate (HR) from 31.31% to 2.44%, a 92.21% relative reduction, while gated projection reduces HR to 3.74%, an 88.05% relative reduction, with lower false rejection of genuine speech. On LibriSpeech, gated projection increases absolute word error rate (WER) by 0.33-4.39 percentage points and yields false-rejection rates (FRR) of 0.41--9.97% across model and split settings. These results show that low-rank activation projection can substantially suppress Whisper hallucinations without retraining, while providing a controllable trade-off between hallucination suppression and speech recognition performance.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-07 in cs.AI).
Categorie arXiv: Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2609.04561 — 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 2026-10-16T22:18:11.363Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-07?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).