Imagen de Portada

Do We Need Distinct Representations for Every Speech Token? Unveiling and Exploiting Redundancy in Large Speech Language Models

Fuente: arXiv

Guardado en:

Detalles Bibliográficos
Autores principales:	Xiang, Bajian, Guo, Tingwei, Chen, Xuan, Han, Yang
Formato:	Preprint
Publicado:	2026
Materias:	Computation and Language Artificial Intelligence
Acceso en línea:	Acceder al recurso
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

Ejemplares similares

Understanding the Modality Gap: An Empirical Study on the Speech-Text Alignment Mechanism of Large Speech Language Models
por: Xiang, Bajian, et al.
Publicado: (2025)

Advancing Speech Language Models by Scaling Supervised Fine-Tuning with Over 60,000 Hours of Synthetic Speech Dialogue Data
por: Zhao, Shuaijiang, et al.
Publicado: (2024)

SpeechR: A Benchmark for Speech Reasoning in Large Audio-Language Models
por: Yang, Wanqi, et al.
Publicado: (2025)

Every Token Counts: Generalizing 16M Ultra-Long Context in Large Language Models
por: Hu, Xiang, et al.
Publicado: (2025)

Measuring the Redundancy of Decoder Layers in SpeechLLMs
por: Moumen, Adel, et al.
Publicado: (2026)

TS-PEFT: Unveiling Token-Level Redundancy in Parameter-Efficient Fine-Tuning
por: Ma, Dabiao, et al.
Publicado: (2025)

VITA-Audio: Fast Interleaved Cross-Modal Token Generation for Efficient Large Speech-Language Model
por: Long, Zuwei, et al.
Publicado: (2025)

LoASR-Bench: Evaluating Large Speech Language Models on Low-Resource Automatic Speech Recognition Across Language Families
por: Chen, Jianan, et al.
Publicado: (2026)

Unveiling Selection Biases: Exploring Order and Token Sensitivity in Large Language Models
por: Wei, Sheng-Lun, et al.
Publicado: (2024)

MOSS-Speech: Towards True Speech-to-Speech Models Without Text Guidance
por: Zhao, Xingjian, et al.
Publicado: (2025)

SpeechMedAssist: Efficiently and Effectively Adapting Speech Language Models for Medical Consultation
por: Chen, Sirry, et al.
Publicado: (2026)

DM-Codec: Distilling Multimodal Representations for Speech Tokenization
por: Ahasan, Md Mubtasim, et al.
Publicado: (2024)

Preservation of Language Understanding Capabilities in Speech-aware Large Language Models
por: Kubis, Marek, et al.
Publicado: (2025)

Cross-Modal Knowledge Distillation for Speech Large Language Models
por: Wang, Enzhi, et al.
Publicado: (2025)

Cross-Lingual Conversational Speech Summarization with Large Language Models
por: Nelson, Max, et al.
Publicado: (2024)

Data Quality Issues in Multilingual Speech Datasets: The Need for Sociolinguistic Awareness and Proactive Language Planning
por: Lau, Mingfei, et al.
Publicado: (2025)

SPBA: Utilizing Speech Large Language Model for Backdoor Attacks on Speech Classification Models
por: Yao, Wenhan, et al.
Publicado: (2025)

Counting Ability of Large Language Models and Impact of Tokenization
por: Zhang, Xiang, et al.
Publicado: (2024)

findsylls: A Language-Agnostic Toolkit for Syllable-Level Speech Tokenization and Embedding
por: Martínez, Héctor Javier Vázquez
Publicado: (2026)

Do Large Language Models Need a Content Delivery Network?
por: Cheng, Yihua, et al.
Publicado: (2024)

InfiniSST: Simultaneous Translation of Unbounded Speech with Large Language Model
por: Ouyang, Siqi, et al.
Publicado: (2025)

FastAdaSP: Multitask-Adapted Efficient Inference for Large Speech Language Model
por: Lu, Yichen, et al.
Publicado: (2024)

Do We Need Frontier Models to Verify Mathematical Proofs?
por: Naik, Aaditya, et al.
Publicado: (2026)

Large Language Models are Efficient Learners of Noise-Robust Speech Recognition
por: Hu, Yuchen, et al.
Publicado: (2024)

When Every Token Counts: Optimal Segmentation for Low-Resource Language Models
por: Raj, Bharath, et al.
Publicado: (2024)

UITron-Speech: Towards Automated GUI Agents Based on Speech Instructions
por: Han, Wenkang, et al.
Publicado: (2025)

SpeechPrune: Context-aware Token Pruning for Speech Information Retrieval
por: Lin, Yueqian, et al.
Publicado: (2024)

TIDE: Every Layer Knows the Token Beneath the Context
por: Jaiswal, Ajay, et al.
Publicado: (2026)

Conditioning Large Language Models on Legal Systems? Detecting Punishable Hate Speech
por: Ludwig, Florian, et al.
Publicado: (2025)

Locate-and-Focus: Enhancing Terminology Translation in Speech Language Models
por: Wu, Suhang, et al.
Publicado: (2025)

Streamlining Redundant Layers to Compress Large Language Models
por: Chen, Xiaodong, et al.
Publicado: (2024)

GenTranslate: Large Language Models are Generative Multilingual Speech and Machine Translators
por: Hu, Yuchen, et al.
Publicado: (2024)

From Flat Language Labels to Typological Priors: Structured Language Conditioning for Multilingual Speech-to-Speech Translation
por: Pan, Yu, et al.
Publicado: (2026)

Speech-FT: Merging Pre-trained And Fine-Tuned Speech Representation Models For Cross-Task Generalization
por: Lin, Tzu-Quan, et al.
Publicado: (2025)

Cross-Lingual Interleaving for Speech Language Models
por: Moumen, Adel, et al.
Publicado: (2025)

Efficient Training for Cross-lingual Speech Language Models
por: Zhou, Yan, et al.
Publicado: (2026)

Enhancing Large Language Models for Mobility Analytics with Semantic Location Tokenization
por: Chen, Yile, et al.
Publicado: (2025)

First Token Probability Guided RAG for Telecom Question Answering
por: Chen, Tingwei, et al.
Publicado: (2025)

Incorporating Contextual Paralinguistic Understanding in Large Speech-Language Models
por: Wang, Qiongqiong, et al.
Publicado: (2025)

SpeechIQ: Speech-Agentic Intelligence Quotient Across Cognitive Levels in Voice Understanding by Large Language Models
por: Wan, Zhen, et al.
Publicado: (2025)