This thesis shows that hidden-state analysis is a rich tool for examining how neural ASR models organize ambiguous speech, and highlights the limits of directly mapping human psycholinguistic designs onto ASR systems.
The degradation rate across neural models, both sentence embeddings and decoder-only LLMs, is studied, and how consistent it is depends on the scale of the noise: under word-level noise, models with very different architectures decline along nearly the same curve, while under character-level noise they separate.
The results show that models’ verb representations degrade more when the authors perturb word order, which removes syntactic information, than when they perturb co-occurrence among content words, which removes semantic information.
Xiao-Meng Zhu, R. Thomas McCoy, Robert Frank· Open Mind· 0 citations
Sparse AutoEncoders (SAEs) offer a promising way to inspect language model representations, but it is still unclear what kind of linguistic structure their latents expose. We use part-of-speech (PoS) categories as a controlled test case to study whether morpho-syntactic information is encoded by individual latents or b...
Alessandro Bondielli, Lucia C. Passaro, Serena Auriemma et al.· 0 citations
The results show that encoder-based metrics remain highly competitive, while generative LLMs perform strongly in hypothesis comparison and improve the interpretability of ASR evaluation.
Thibault Bañeras-Roux, Shashi Kumar, Driss Khalil et al.· 0 citations
Text-to-speech models for Bantu tonal languages are challenged by a tonal system that is rooted in both the lexis (i.e., the inventory of words, stems, and affixes) and the grammar (i.e., morpho-syntax). To complicate matters, the standard writing systems of these languages often omit tone markings and syllable duratio...
It is demonstrated that valence-related activation patterns leave hidden traces that predictably govern later choices, even when all visible tokens are identical across conditions, and whether these traces are accompanied by any subjective experience relevant to model welfare remains unclear.
Cameron Berg, Caspar Kaiser· 1 citation· ⚡1
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
