Publications (26)
Generic behavior of differentially positive systems on a globally orderable Riemannian manifold
Lin Niu, Yi Wang
Differentially positive systems are the nonlinear systems whose linearization along trajectories preserves a cone field on a smooth Riemannian manifold. One of the embryonic forms…
Improving Post-Training Quantization on Object Detection with Task Loss-Guided Lp Metric
Lin Niu, Jiawei Liu, Zhihang Yuan +3
Efficient inference for object detection networks is a major challenge on edge devices. Post-Training Quantization (PTQ), which transforms a full-precision model into low bit-width…
PD-Quant: Post-Training Quantization based on Prediction Difference Metric
Jiawei Liu, Lin Niu, Zhihang Yuan +3
Post-training quantization (PTQ) is a neural network compression technique that converts a full-precision model into a quantized model using lower-precision data types. Although it…
Generic Poincaré-Bendixson Theorem for singularly perturbed monotone systems with respect to cones of rank-
Lin Niu, Xizhuang Xie
We investigate the singularly perturbed monotone systems with respect to cones of rank and obtain the so called Generic Poincaré-Bendixson theorem for such perturbed systems,…
CoSA: Accelerating Long-Context Inference via Proxy-Kernel Co-Designed Sparse Attention
Yufei Xue, Lin Niu, Hong Liu +6
CoSA introduces a training-free, two-stage sparse attention method that jointly designs a proxy and kernel to efficiently handle very long contexts, achieving faster inference with…
Solitary wave structure of transitional flow in the wake of a sphere
Lin Niu, Hua-Shu Dou, Changquan Zhou +1
The soliton-like coherent structure (SCS), which has been verified to exist in both transitional and turbulent boundary layers1-4, still poses a challenge in the understanding of i…
Stem: Rethinking Causal Information Flow in Sparse Attention
Lin Niu, Xin Luo, Linchuan Xie +4
The quadratic computational complexity of self-attention remains a fundamental bottleneck for scaling Large Language Models (LLMs) to long contexts, particularly during the pre-fil…
Large-area and high-quality 2D transition metal telluride
Jiadong Zhou, Fucai Liu, Junhao Lin +24
Atomically thin transitional metal ditellurides like WTe2 and MoTe2 have triggered tremendous research interests because of their intrinsic nontrivial band structure. They are also…
HunyuanVideo 1.5 Technical Report
Bing Wu, Chang Zou, Changlin Li +78
We present HunyuanVideo 1.5, a lightweight yet powerful open-source video generation model that achieves state-of-the-art visual quality and motion coherence with only 8.3 billion…
E-Sparse: Boosting the Large Language Model Inference through Entropy-based N:M Sparsity
Yun Li, Lin Niu, Xipeng Zhang +3
Traditional pruning methods are known to be challenging to work in Large Language Models (LLMs) for Generative AI because of their unaffordable training process and large computati…
Hunyuan3D 2.0: Scaling Diffusion Models for High Resolution Textured 3D Assets Generation
Zibo Zhao, Zeqiang Lai, Qingxiang Lin +71
We present Hunyuan3D 2.0, an advanced large-scale 3D synthesis system for generating high-resolution textured 3D assets. This system includes two foundation components: a large-sca…
Hunyuan3D Studio: End-to-End AI Pipeline for Game-Ready 3D Asset Generation
Biwen Lei, Yang Li, Xinhai Liu +97
The creation of high-quality 3D assets, a cornerstone of modern game development, has long been characterized by labor-intensive and specialized workflows. This paper presents Huny…
RPTQ: Reorder-based Post-training Quantization for Large Language Models
Zhihang Yuan, Lin Niu, Jiawei Liu +7
Large-scale language models (LLMs) have demonstrated impressive performance, but their deployment presents challenges due to their significant memory usage. This issue can be allev…
PIVOT: Efficient Query-Group Indexing for Token-Level Sparse Attention
Hong Liu, Yuan Cheng, Lin Niu +5
Token-level sparse attention, as implemented by DeepSeek Sparse Attention (DSA) in production systems, makes the downstream attention efficient but shifts the bottleneck to the ind…
Sharper Ramsey lower bounds from refined Gaussian estimates
Qizhong Lin, Lin Niu
Recently, Ma, Shen and Xie broke the ErdÅs barrier for off-diagonal Ramsey numbers , achieving the first exponential improvement over the classical lower bound for…
AngelSlim: A more accessible, comprehensive, and efficient toolkit for large model compression
Rui Cen, QiangQiang Hu, Hong Huang +10
This technical report introduces AngelSlim, a comprehensive and versatile toolkit for large model compression developed by the Tencent Hunyuan team. By consolidating cutting-edge a…
Non-Oscillation Principle for Eventually Competitive and Cooperative Systems
Lin Niu, Yi Wang
A nonlinear dynamical system is called eventually competitive (or cooperative) provided that it preserves a partial order in backward (or forward) time only after some reasonable i…
Controlled Synthesis of Organic/Inorganic van der Waals Solid for Tunable Light-matter Interactions
Lin Niu, Xinfeng Liu, Chunxiao Cong +22
Van der Waals (vdW) solids, as a new type of artificial materials that consist of alternating layers bonded by weak interactions, have shed light on fascinating optoelectronic devi…
Effect of gap width on turbulent transition in Taylor-Couette flow
Chang-Quan Zhou, Hua-Shu Dou, Lin Niu +1
Simulations of the transitional flow in Taylor-Couette configuration are carried out to study the effect of the gap width on turbulent transition. The research results show that, u…
Heteroclinic Cycles of A Symmetric May-Leonard Competition Model with Seasonal Succession
Xizhuang Xie, Lin Niu
In this paper, we are concerned with the stability of heteroclinic cycles of the symmetric May-Leonard competition model with seasonal succession. Sufficient conditions for stabili…
Hypergraph ErdÅs--Rogers functions with consecutive clique sizes
Qizhong Lin, Lin Niu
For integers \(k\le s<t\), the hypergraph ErdÅs--Rogers function \(f^{(k)}_{s,t}(n)\) is the largest integer \(m\) such that every \(n\)-vertex \(K_t^{(k)}\)-free \(k\)-graph cont…
Almost sure convergence of differentially positive systems on a globally orderable manifold
Lin Niu, Yi Wang, Yufeng Zhang
Differentially positive systems are nonlinear systems whose linearization along trajectories preserves a cone field on a smooth Riemannian manifold. The structures of cone field co…
Decompositions of graphs of nonnegative characteristic with some forbidden subgraphs
Lin Niu, Xiangwen Li
A {\em -decomposition} of a graph is an order pair such that is a subgraph of where has the maximum degree at most and is an acyclic orientat…
Inverse Energy Cascade in Turbulent Taylor-Couette Flows
Changquan Zhou, Hua-Shu Dou, Lin Niu +1
The inverse energy cascade in turbulent Taylor-Couette flow is studied in line with the results of the large eddy simulation. The simulation results show that the inverse energy ca…
HunyuanWorld 1.0: Generating Immersive, Explorable, and Interactive 3D Worlds from Words or Pixels
HunyuanWorld Team, Zhenwei Wang, Yuhao Liu +52
Creating immersive and playable 3D worlds from texts or images remains a fundamental challenge in computer vision and graphics. Existing world generation approaches typically fall…
Hunyuan3D 2.1: From Images to High-Fidelity 3D Assets with Production-Ready PBR Material
Team Hunyuan3D, Shuhui Yang, Mingxin Yang +50
3D AI-generated content (AIGC) is a passionate field that has significantly accelerated the creation of 3D models in gaming, film, and design. Despite the development of several gr…