Researcher — paper.dinq.me

2026 顶会论文

一作ICLR 2026Poster· Deep Learning / Generative Models and Autoencoders

Tracing the Principles Behind Modern Diffusion Models

Chieh-Hsin Lai · Yang Song · Dongjun Kim · Yuki Mitsufuji · Stefano Ermon

OpenReview

作者位次 #2/4ICLR 2026Poster· Computer Vision / Image and Video Generation

CMT: Mid-Training for Efficient Learning of Consistency, Mean Flow, and Flow-Map Models

Zheyuan Hu · Chieh-Hsin Lai · Yuki Mitsufuji · Stefano Ermon

多模态效率与系统OpenReview

作者位次 #8/9ICLR 2026Poster

SONA: Learning Conditional, Unconditional, and Matching-Aware Discriminator

Yuhta Takida · Satoshi Hayakawa · Takashi Shibuya · Masaaki Imaizumi · Naoki Murata · Bac Nguyen · Toshimitsu Uesaka · Chieh-Hsin Lai · Yuki Mitsufuji

OpenReview

作者位次 #3/6ICLR 2026Poster· Deep Learning / Generative Models and Autoencoders

Improving Classifier-Free Guidance in Masked Diffusion: Low-Dim Theoretical Insights with High-Dim Impact

Kevin Rojas · Ye He · Chieh-Hsin Lai · Yuhta Takida · Yuki Mitsufuji · Molei Tao

理论与学习动力学OpenReview

作者位次 #4/7ICLR 2026Poster· Deep Learning / Generative Models and Autoencoders

Improved Object-Centric Diffusion Learning with Registers and Contrastive Alignment

Bac Nguyen · Yuhta Takida · Naoki Murata · Chieh-Hsin Lai · Toshimitsu Uesaka · Stefano Ermon · Yuki Mitsufuji

对齐与偏好OpenReview

作者位次 #2/10ICLR 2026Poster· Computer Vision / Image and Video Generation

Concept-TRAK: Understanding how diffusion models learn concepts through concept attribution

Yong-Hyun Park · Chieh-Hsin Lai · Satoshi Hayakawa · Yuhta Takida · Naoki Murata · WeiHsiang Liao · Woosung Choi · Kin Wai Cheuk · Junghyun (Tony) Koo · Yuki Mitsufuji

多模态OpenReview

作者位次 #2/5CVPR 2026Poster· Computer Vision & Pattern Recognition

MeanFlow Transformers with Representation Autoencoders

Zheyuan Hu · Chieh-Hsin Lai · Ge Wu · Yuki Mitsufuji · Stefano Ermon

可解释性多模态OpenReview 项目页

作者位次 #5/5KDD 2026Oral· Knowledge Discovery & Data Mining

Stabilizing Physics-Informed Consistency Models via Structure-Preserving Training

Che-Chia Chang · Chen-Yang Dai · Te-Sheng Lin · Ming-Chih Lai · Chieh-Hsin Lai

检索与知识OpenReview

作者位次 #5/11AAAI 2026Poster· Artificial Intelligence

SteerMusic: Enhanced Musical Consistency for Zero-shot Text-Guided and Personalized Music Editing

Xinlei Niu · Kin Wai Cheuk · Jing Zhang · Naoki Murata · Chieh-Hsin Lai · Michele Mancusi · Woosung Choi · Giorgio Fabbro · Wei-Hsiang Liao · Charles Patrick Martin · Yuki Mitsufuji