The paper investigates how Spoken Language Models (SLMs) process speech compared to text, noting that current SLMs show weak alignment between speech and text representations despite strong downstream performance. The authors propose a framework that separates length mismatch from semantic alignment to better match speech and text representations. Experiments on multiple benchmarks demonstrate that this approach yields competitive results against strong baselines, highlighting the need to explicitly address structural differences between speech and text in SLM training.
By Hyeonyu Kim, Hwayeon Kim, Youngwon Choi, Myeongkyun Cho, Huu-Kim Nguyen
SPAR-K is a scheduled periodic alternating early‑exit framework for interleaved spoken language models that reduces decoding depth for speech tokens while maintaining quality. It lets most speech positions exit at a fixed intermediate layer and inserts periodic full‑depth refresh steps to counter distribution shift. Experiments on Step‑Audio‑2‑mini and GLM‑4‑Voice show up to 11 % depth reduction with less than 0.82 % drop in question‑answering accuracy and negligible impact on MOS and WER.
By Hsiao-Ying Huang, Cheng-Han Chiang, Hung-yi Lee
Connecting a pre-trained speech encoder to a Large Language Model (LLM) is the standard architecture for building Speech LLMs. However, a structural misalignment exists between the encoder and the LLM.
arXiv:2506. 16738v2 Announce Type: replace-cross Abstract: With the rapid progress of speech language models (SLMs), discrete speech tokens have emerged as a core interface between speech and text, enabling unified modeling across modalities.
By Daejin Jo, Jeeyoung Yun, Byungseok Roh, Sungwoong Kim
arXiv:2606. 10029v1 Announce Type: cross Abstract: Language models increasingly serve as the backbone of text-to-speech (TTS) systems, yet we understand little about the representations they build when text and generated speech tokens share a single residual stream.
By Nikita Koriagin, Georgii Aparin, Nikita Balagansky, Daniil Gavrilov
arXiv:2609.10434v1 Announce Type: new
Abstract: Self-supervised speech foundation models are now used in a wide array of downstream applications, including traditional speech recognition and as the b...
By Robin Huo, Ewan Dunbar
The paper introduces Hybrid Search, a method that refines warm-initialized large language model (LLM) based automatic speech recognition (ASR) systems by exploiting interactions between ASR hidden states and the base LLM’s hidden states. By identifying tokens with high semantic dependence and selectively correcting them, the approach surpasses traditional global LLM‑correction techniques such as rescoring and late fusion. The study demonstrates that even after warm initialization, LLM‑based ASR models can further benefit from their base LLM during inference.
By Chan-Jan Hsu, Jaeyeon Kim, Chao-Han Huck Yang, Shinji Watanabe, Hung-yi Lee, Carlos Busso
arXiv:2609.09719v1 Announce Type: new
Abstract: Text-aligned speech tokenization methods have emerged to better align speech tokens with LLM token spaces, enabling more effective utilization of pretr...
By Kang-wook Kim, Jinyoung Park, Jinsoo Kim, Sehun Lee, Sang Hoon Woo, Gunhee Kim
arXiv:2509.10452v3 Announce Type: replace-cross
Abstract: Pretrained automatic speech recognition (ASR) models such as Whisper perform well but still need domain adaptation to handle unseen parlance....
By Akshat Pandey, Karun Kumar, Raphael Tang
Text-aligned speech tokenization methods have emerged to better align speech tokens with LLM token spaces, enabling more effective utilization of pretrained LLMs. However, they rely on offline automat...
arXiv:2605.28456v2 Announce Type: replace
Abstract: Existing Visual Speech Recognition (VSR) systems commonly rely on left-to-right autoregressive decoding, which can force premature decisions on vis...
By Jeong Hun Yeo, Chae Won Kim, Hyeongseop Rha, Yong Man Ro
The study investigates how textual neural models degrade when inputs contain noise such as typos, OCR errors, or dropped words. It finds that model performance decline is largely consistent across architectures under word‑level noise but diverges under character‑level noise, a difference attributed to tokenization rather than architecture. By applying a short contrastive training recipe, diverse encoders converge to a common robustness curve, enabling prediction of a model’s noise resilience and the ability to enhance robustness at specific noise scales through targeted training.
By Yefan Tao, Gerald Friedland, Luyang Kong