3 citations · 7 across the 9 of their papers we have counts for
12 papers
Beyond Prediction: A Component-Level Study of Concept Recovery in In-Context Learning
Xufeng Liu, Fang Wan, Yi Liu
In-context learning (ICL) enables language models to make accurate predictions from demonstrations, but prediction success alone does not reveal the extent to which the underlying…
Sparse Attention Is Matrix Approximation, Not Choosing from a Bag of Values
Fang Wan, Xufeng Liu, Fan Li +1
Large Language Models (LLMs) achieve strong performance across many domains, but their efficiency is limited by the quadratic cost of attention with respect to prompt length. Spars…
ActionUNet: Improving Robustness of VLA Models with Efficient Multi-scale Fine-tuning
Di Zhu, Ziheng Yan, Fang Wan
Vision-Language-Action (VLA) models have shown great promise for robotic manipulation by mapping multi-modal semantics to physical actions. However, this mapping inherently struggl…
Vorch-Omni: Multi-Task Orchestration of Sight and Sound
Vorch Team, Xiaoyu Chen, Yang Ding +25
Recent advances in generative video modeling have enabled diverse generation, reference-based synthesis, extension, and editing, but existing approaches often rely on fragmented ta…
Thinking with Images via Self-Calling Agent
Wenxi Yang, Yuzhong Zhao, Fang Wan +1
Thinking-with-images paradigms have showcased remarkable visual reasoning capability by integrating visual information as dynamic elements into the Chain-of-Thought (CoT). However,…
Geometric-Mean Policy Optimization
Yuzhong Zhao, Yue Liu, Junpeng Liu +9
Group Relative Policy Optimization (GRPO) has significantly enhanced the reasoning capability of large language models by optimizing the arithmetic mean of token-level rewards. Unf…
Timestep Embedding Tells: It's Time to Cache for Video Diffusion Model
Feng Liu, Shiwei Zhang, Xiaofeng Wang +6
As a fundamental backbone for video generation, diffusion models are challenged by low inference speed due to the sequential nature of denoising. Previous methods speed up the mode…
Correspondence-Guided SfM-Free 3D Gaussian Splatting for NVS
Wei Sun, Xiaosong Zhang, Fang Wan +4
Novel View Synthesis (NVS) without Structure-from-Motion (SfM) pre-processed camera poses--referred to as SfM-free methods--is crucial for promoting rapid response capabilities and…
Evaluation of Text-to-Video Generation Models: A Dynamics Perspective
Mingxiang Liao, Hannan Lu, Xinyu Zhang +6
Comprehensive and constructive evaluation protocols play an important role in the development of sophisticated text-to-video (T2V) generation models. Existing evaluation protocols…
DynRefer: Delving into Region-level Multimodal Tasks via Dynamic Resolution
Yuzhong Zhao, Feng Liu, Yue Liu +4
One fundamental task of multimodal models is to translate referred image regions to human preferred language descriptions. Existing methods, however, ignore the resolution adaptabi…
Ray Denoising: Depth-aware Hard Negative Sampling for Multi-view 3D Object Detection
Feng Liu, Tengteng Huang, Qianjing Zhang +5
Multi-view 3D object detection systems often struggle with generating precise predictions due to the challenges in estimating depth from images, increasing redundant and incorrect…
ControlCap: Controllable Region-level Captioning
Yuzhong Zhao, Yue Liu, Zonghao Guo +4
Region-level captioning is challenged by the caption degeneration issue, which refers to that pre-trained multimodal models tend to predict the most frequent captions but miss the…