影响力指数

55.28/100

前 6.3%

全站排名 #4,084

发表论文15 篇

平均评分5.4

年均产出5.0 篇/年

Boran Han

Researcher@Amazon/AWS·美国·OpenReview

研究方向

representation learning · computer vision · remote sensing · miscoscopy · physics-based model · foundation models · multimodal foundation models · in-context learning · unsupervised pretraining · AI for bioscience · AI for geoscience · LLM

6.5

Understanding Transformers for Time Series: Rank Structure, Flow-of-ranks, and Compressibility

ICLR 2026Poster

三作

5.5

Understanding the Implicit Biases of Design Choices for Time Series Foundation Models

ICLR 2026Poster

三作

3.5

Enhancing Foundation Models for Time Series Forecasting via Wavelet-based Tokenization

ICML 2025Poster

三作

5.5

Effectively Steer LLM To Follow Preference via Building Confident Directions

ICLR 2025Rejected

二作

5.5

Enhancing Foundation Models for Time Series Forecasting via Wavelet-based Tokenization

ICLR 2025Rejected

三作

5.4

Gradient-Free Generation for Hard-Constrained Systems

ICLR 2025Poster

二作

4.8

Comparing and Contrasting Deep Learning Weather Prediction Backbones on Navier-Stokes and Atmospheric Dynamics

ICLR 2025Rejected

3.8

Forecasting Needles in a Time Series Haystack

合作者 (20)

Boran Han

Understanding Transformers for Time Series: Rank Structure, Flow-of-ranks, and Compressibility

Understanding the Implicit Biases of Design Choices for Time Series Foundation Models

When Does Multimodality Lead to Better Time Series Forecasting?

Mitra: Mixed Synthetic Priors for Enhancing Tabular Foundation Models

MLZero: A Multi-Agent System for End-to-end Machine Learning Automation

Enhancing Foundation Models for Time Series Forecasting via Wavelet-based Tokenization

Effectively Steer LLM To Follow Preference via Building Confident Directions

Enhancing Foundation Models for Time Series Forecasting via Wavelet-based Tokenization

Gradient-Free Generation for Hard-Constrained Systems

Comparing and Contrasting Deep Learning Weather Prediction Backbones on Navier-Stokes and Atmospheric Dynamics

Forecasting Needles in a Time Series Haystack