PaperHub
影响力指数
80.89/100
前 1.3%
全站排名 #835
发表论文25 篇
平均评分5.3
年均产出8.3 篇/年

Zeke Xie

Assistant Professor@HKUST(GZ)·中国·OpenReview
研究方向

Large Language Models · Generative AI · Diffusion Models · Deep Learning Theory · Optimization · Machine Learning

5.6
27

Weak-to-Strong Diffusion with Reflection

ICLR 2026Poster
三作
5.5
18

Multiphysics Bench: Benchmarking and Investigating Scientific Machine Learning for Multiphysics PDEs

ICLR 2026Rejected
通讯
5.5
21

Late-to-Early Training: LET LLMs Learn Earlier, So Faster and Better

ICLR 2026Poster
通讯
5.5
11

CMPhysBench: A Benchmark for Evaluating Large Language Models in Condensed Matter Physics

ICLR 2026Poster
5.0
16

Guidance Matters: Rethinking the Evaluation Pitfall for Text-to-Image Generation

ICLR 2026Poster
通讯
4.5
28

Reflective Flow Sampling Enhancement

ICLR 2026Rejected
通讯
4.0
17

Dynamic Locally Linear Graph Learning for Geometry-Aware GNNs

ICLR 2026Desk Rejected
三作
4.0
5

Do All Individual Layers Help? An Empirical Study of Task-Interfering Layers in Vision-Language Models

ICLR 2026Withdrawn
3.6
6

Diffusion Dataset Condensation: Training Your Diffusion Model Faster with Less Data

ICLR 2026Withdrawn
通讯
3.2
18

Principal Spectral Regularization Makes Momentum Surpass Adam for LLM Training

ICLR 2026Rejected
通讯
8.2
22

Channel Matters: Estimating Channel Influence for Multivariate Time Series

NeurIPS 2025Poster
二作
6.8
17

UtilGen: Utility-Centric Generative Data Augmentation with Dual-Level Task Adaptation

NeurIPS 2025Poster
6.4
32

MAE-Pure: Semantic-Preserving Adversarial Purification

NeurIPS 2025Rejected
6.3
17

IV-mixed Sampler: Leveraging Image Diffusion Models for Enhanced Video Synthesis

ICLR 2025Poster
通讯
6.0
22

Diff-In: Data Influence Estimation with Differential Approximation

ICLR 2025Rejected
6.0
25

Zigzag Diffusion Sampling: Diffusion Models Can Self-Improve via Self-Reflection

ICLR 2025Poster
通讯
5.5
35

Channel-wise Influence: Estimating Data Influence for Multivariate Time Series

ICLR 2025Withdrawn
二作
5.5
13

Principled Data Selection for Alignment: The Hidden Risks of Difficult Examples

ICML 2025Poster
4.8
7

Noise Prompt Learning: Learning the Winning Tickets for Diffusion Sampling

ICLR 2025Withdrawn
通讯
4.8
9

Investigating the Overlooked Hessian Structure: From CNNs to LLMs

ICML 2025Poster
通讯
4.5
5

Not All Noises Are Created Equally: Diffusion Noise Selection and Optimization

ICLR 2025Withdrawn
通讯
4.3
6

The Blessing of Smooth Initialization for Video Diffusion Models

ICLR 2025Withdrawn
通讯
合作者 (20)