arXiv — NLP / Computation & Language · · 3 min read

Transcription Policy as a Latent Variable: Activating Controllable Verbatim ASR with Word-Level Timing

Mirrored from arXiv — NLP / Computation & Language for archival readability. Support the source by reading on the original site.

Computer Science > Computation and Language

arXiv:2607.18934 (cs)
[Submitted on 21 Jul 2026]

Title:Transcription Policy as a Latent Variable: Activating Controllable Verbatim ASR with Word-Level Timing

Authors:Laurin Wagner (1), Mario Zusag (1), Bernhard Thallinger (1) ((1) nyra labs)
View a PDF of the paper titled Transcription Policy as a Latent Variable: Activating Controllable Verbatim ASR with Word-Level Timing, by Laurin Wagner (1) and 2 other authors
View PDF HTML (experimental)
Abstract:Modern ASR models trained on heterogeneously annotated data treat transcription style (verbatim vs. intended) as an uncontrolled latent variable, causing measurable decoding instability, evaluation confounding (up to 60% of reported WER attributable to style mismatch), and unreliable word-level timing. We show that models already encode both styles; the challenge is controlled activation. Using coverage-aware decoder task tokens trained on parallel verbatim/intended transcript pairs, we raise German disfluency F1 from 10% to 79% zero-shot, despite English-only training. Full English-only fine-tuning surpasses all baselines in verbatim accuracy, disfluency detection, and intended-mode quality across both languages. We further introduce supervised cross-attention fine-tuning that improves word-level timestamps on disfluent speech beyond forced-alignment baselines. Finally, we propose verbatimize, a new task enabling scalable creation and enrichment of speech corpora with high-quality canonical verbatim transcriptions.
Comments: Accepted at Interspeech 2026 long track
Subjects: Computation and Language (cs.CL)
Cite as: arXiv:2607.18934 [cs.CL]
  (or arXiv:2607.18934v1 [cs.CL] for this version)
  https://doi.org/10.48550/arXiv.2607.18934
arXiv-issued DOI via DataCite (pending registration)

Submission history

From: Mario Zusag [view email]
[v1] Tue, 21 Jul 2026 10:19:17 UTC (216 KB)
Full-text links:

Access Paper:

    View a PDF of the paper titled Transcription Policy as a Latent Variable: Activating Controllable Verbatim ASR with Word-Level Timing, by Laurin Wagner (1) and 2 other authors
  • View PDF
  • HTML (experimental)
  • TeX Source

Current browse context:

cs.CL
< prev   |   next >
Change to browse by:
cs

References & Citations

Loading...

BibTeX formatted citation

loading...
Data provided by:

Bookmark

BibSonomy Reddit
Bibliographic Tools

Bibliographic and Citation Tools

Bibliographic Explorer Toggle
Bibliographic Explorer (What is the Explorer?)
Connected Papers Toggle
Connected Papers (What is Connected Papers?)
Litmaps Toggle
Litmaps (What is Litmaps?)
scite.ai Toggle
scite Smart Citations (What are Smart Citations?)
Code, Data, Media

Code, Data and Media Associated with this Article

alphaXiv Toggle
alphaXiv (What is alphaXiv?)
Links to Code Toggle
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
DagsHub Toggle
DagsHub (What is DagsHub?)
GotitPub Toggle
Gotit.pub (What is GotitPub?)
Huggingface Toggle
Hugging Face (What is Huggingface?)
ScienceCast Toggle
ScienceCast (What is ScienceCast?)
Demos

Demos

Replicate Toggle
Replicate (What is Replicate?)
Spaces Toggle
Hugging Face Spaces (What is Spaces?)
Spaces Toggle
TXYZ.AI (What is TXYZ.AI?)
Related Papers

Recommenders and Search Tools

Link to Influence Flower
Influence Flower (What are Influence Flowers?)
Core recommender toggle
CORE Recommender (What is CORE?)
About arXivLabs

arXivLabs: experimental projects with community collaborators

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.

Discussion (0)

Sign in to join the discussion. Free account, 30 seconds — email code or GitHub.

Sign in →

No comments yet. Sign in and be the first to say something.

More from arXiv — NLP / Computation & Language