6 papers
Learning Native Continuation for Action Chunking Flow Policies
Yufeng Liu, Hang Yu, Juntu Zhao +9
Action chunking enables Vision Language Action (VLA) models to run in real time, but naive chunked execution often exhibits discontinuities at chunk boundaries. Real-Time Chunking…
EVA01: Unified Native 3D Understanding and Generation via Mixture-of-Transformers
Zongyuan Yang, Mingjing Yi, Wanli Ma +8
This paper addresses the challenge of integrating 3D meshes as a native modality within Multimodal Large Language Models (MLLMs). Diffusion-based large reconstruction models decoup…
Input Domain Aware MoE: Decoupling Routing Decisions from Task Optimization in Mixture of Experts
Yongxiang Hua, Haoyu Cao, Zhou Tao +4
Sparse Mixture of Experts (sMoE) has become a pivotal approach for scaling large vision-language models, offering substantial capacity while maintaining computational efficiency th…
Addressing Representation Collapse in Vector Quantized Models with One Linear Layer
Yongxin Zhu, Bocheng Li, Yifei Xin +2
Vector Quantization (VQ) is essential for discretizing continuous representations in unsupervised learning but suffers from representation collapse, causing low codebook utilizatio…
Unifying Continuous and Discrete Text Diffusion with Non-simultaneous Diffusion Processes
Bocheng Li, Zhujin Gao, Linli Xu
Diffusion models have emerged as a promising approach for text generation, with recent works falling into two main categories: discrete and continuous diffusion models. Discrete di…
Stabilize the Latent Space for Image Autoregressive Modeling: A Unified Perspective
Yongxin Zhu, Bocheng Li, Hang Zhang +3
Latent-based image generative models, such as Latent Diffusion Models (LDMs) and Mask Image Models (MIMs), have achieved notable success in image generation tasks. These models typ…