10 papers · 1 filter
Morphing into Hybrid Attention Models
Disen Lan, Jianbin Zheng, Yuxi Ren +5
Hybrid attention models improve long-context efficiency by retaining only a subset of full-attention layers and replacing the remaining layers with linear attention. However, the e…
Prism: Spectral-Aware Block-Sparse Attention
Xinghao Wang, Pengyu Wang, Xiaoran Liu +4
Block-sparse attention is promising for accelerating long-context LLM pre-filling, yet identifying relevant blocks efficiently remains a bottleneck. Existing methods typically empl…
MOSS-Speech: Towards True Speech-to-Speech Models Without Text Guidance
Xingjian Zhao, Zhe Xu, Qinyuan Cheng +20
Spoken dialogue systems often rely on cascaded pipelines that transcribe, process, and resynthesize speech. While effective, this design discards paralinguistic cues and limits exp…
Sparser Block-Sparse Attention via Token Permutation
Xinghao Wang, Pengyu Wang, Dong Zhang +7
Scaling the context length of large language models (LLMs) offers significant benefits but is computationally expensive. This expense stems primarily from the self-attention mechan…
UnifiedVisual: A Framework for Constructing Unified Vision-Language Datasets
Pengyu Wang, Shaojun Zhou, Chenkun Tan +7
Unified vision large language models (VLLMs) have recently achieved impressive advancements in both multimodal understanding and generation, powering applications such as visual qu…
Decoupled Proxy Alignment: Mitigating Language Prior Conflict for Multimodal Alignment in MLLM
Chenkun Tan, Pengyu Wang, Shaojun Zhou +6
Multimodal large language models (MLLMs) have gained significant attention due to their impressive ability to integrate vision and language modalities. Recent advancements in MLLMs…