Researcher — paper.dinq.me

2026 顶会论文

作者位次 #5/5ICLR 2026Poster· Deep Learning / Generative Models and Autoencoders

SparseD: Sparse Attention for Diffusion Language Models

Zeqing Wang · Gongfan Fang · Xinyin Ma · Xingyi Yang · Xinchao Wang

长上下文与架构OpenReview

作者位次 #4/4ICLR 2026Oral· Social Aspects / Trustworthy Machine Learning

Invisible Safety Threat: Malicious Finetuning for LLM via Steganography

Guangnian Wan · Xinyin Ma · Gongfan Fang · Xinchao Wang

安全社会与人本OpenReview

作者位次 #3/3ICLR 2026Poster

CoDA: From Text-to-Image Diffusion Models to Training-Free Dataset Distillation

Letian Zhou · Songhua Liu · Xinchao Wang

多模态强化学习与后训练数据与预训练OpenReview

作者位次 #5/5ICLR 2026Poster

dParallel: Learnable Parallel Decoding for dLLMs

Zigeng Chen · Gongfan Fang · Xinyin Ma · Ruonan Yu · Xinchao Wang

代码OpenReview

作者位次 #3/3ICLR 2026Poster· Deep Learning / Algorithms

Understanding Dataset Distillation via Spectral Filtering

Deyu Bo · Songhua Liu · Xinchao Wang

强化学习与后训练数据与预训练OpenReview

作者位次 #10/10CVPR 2026Poster· Computer Vision & Pattern Recognition

Rethinking Token Reduction for Large Vision-Language Models

Yi Wang · Haofei Zhang · Qihan Huang · Anda Cao · Gongfan Fang · Wei Wang · Xuan Jin · Jie Song · Mingli Song · Xinchao Wang

多模态OpenReview 项目页

作者位次 #2/2CVPR 2026Poster· Computer Vision & Pattern Recognition

Beyond Soft Label: Dataset Distillation via Orthogonal Gradient Matching

Deyu Bo · Xinchao Wang

多模态强化学习与后训练数据与预训练OpenReview 项目页

作者位次 #5/5CVPR 2026Poster· Computer Vision & Pattern Recognition

Gated Condition Injection without Multimodal Attention: Towards Controllable Linear-Attention Transformers

Yuhe Liu · Zhenxiong Tan · Yujia Hu · Songhua Liu · Xinchao Wang

多模态OpenReview 项目页

作者位次 #6/6CVPR 2026Poster· Computer Vision & Pattern Recognition

Language-guided Frequency Modulation for Large Vision-Language Models

Shuyi Ouyang · Gongfan Fang · Xinyin Ma · Yen-wei Chen · Lanfen Lin · Xinchao Wang

多模态OpenReview 项目页

作者位次 #3/3CVPR 2026Poster· Computer Vision & Pattern Recognition

Merge3D: Efficient 3D Multimodal LLMs via Joint 2D-3D Token Merging

Tianbo Pan · Xingyi Yang · Xinchao Wang

多模态效率与系统OpenReview 项目页

作者位次 #5/5CVPR 2026Poster· Computer Vision & Pattern Recognition

SpotEdit: Selective Region Editing in Diffusion Transformers

Zhibin Qin · Zhenxiong Tan · Zeqing Wang · Songhua Liu · Xinchao Wang

多模态OpenReview 项目页

作者位次 #5/5AAAI 2026Poster· Artificial Intelligence

FreLay: Frequency-aware Energy Function for Training-free Layout-to-Image Generation

Bonan Li · Yinhan Hu · Songhua Liu · Zeyu Xiao · Xinchao Wang

多模态OpenReview

作者位次 #6/6AAAI 2026Poster· Artificial Intelligence

Taming the Phantom: Token-Asymmetric Filtering for Hallucination Mitigation in Large Vision-Language Models

Shuyi Ouyang · Hongyi Wang · Gongfan Fang · Xinyin Ma · Lanfen Lin · Xinchao Wang

安全多模态OpenReview

作者位次 #6/6AAAI 2026Poster· Artificial Intelligence

Minute-Long Videos with Dual Parallelisms

Zeqing Wang · Bowen Zheng · Xingyi Yang · Zhenxiong Tan · Yuecong Xu · Xinchao Wang

多模态OpenReview

作者位次 #2/2AAAI 2026Poster· Artificial Intelligence

Exploiting Blurry Representations for Event-guided Video Super-Resolution

Zeyu Xiao · Xinchao Wang

可解释性多模态OpenReview

Xinchao Wang

获取完整研究档案

联系方式

2026 顶会论文

SparseD: Sparse Attention for Diffusion Language Models

Invisible Safety Threat: Malicious Finetuning for LLM via Steganography

CoDA: From Text-to-Image Diffusion Models to Training-Free Dataset Distillation

dParallel: Learnable Parallel Decoding for dLLMs

Understanding Dataset Distillation via Spectral Filtering

Rethinking Token Reduction for Large Vision-Language Models

Beyond Soft Label: Dataset Distillation via Orthogonal Gradient Matching

Gated Condition Injection without Multimodal Attention: Towards Controllable Linear-Attention Transformers

Language-guided Frequency Modulation for Large Vision-Language Models

Merge3D: Efficient 3D Multimodal LLMs via Joint 2D-3D Token Merging

SpotEdit: Selective Region Editing in Diffusion Transformers

FreLay: Frequency-aware Energy Function for Training-free Layout-to-Image Generation

Taming the Phantom: Token-Asymmetric Filtering for Hallucination Mitigation in Large Vision-Language Models

Minute-Long Videos with Dual Parallelisms

Exploiting Blurry Representations for Event-guided Video Super-Resolution

常见合作者