Preprints
-
A Law of Next-Token Prediction in Large Language Models
Hangfeng He and Weijie Su
In arXiv 2024. [pdf] [code] -
MMCOMPOSITION: Revisiting the Compositionality of Pre-trained Vision-Language Models
Hang Hua, Yunlong Tang, Ziyun Zeng, Liangliang Cao, Zhengyuan Yang, Hangfeng He, Chenliang Xu, and Jiebo Luo
In arXiv 2024. [pdf] -
Rethinking with Retrieval: Faithful Large Language Model Inference
Hangfeng He, Hongming Zhang, and Dan Roth
In arXiv 2023. [pdf] [code]
Publications
2024
-
An Empirical Analysis on Large Language Models in Debate Evaluation
Xinyi Liu, Pinxin Liu, and Hangfeng He
In ACL 2024 (short papers). [pdf] -
SocREval: Large Language Models with the Socratic Method for Reference-Free Reasoning Evaluation
Hangfeng He, Hongming Zhang, and Dan Roth
In NAACL 2024 (findings). [pdf] [code] -
Unveiling Divergent Inductive Biases of LLMs on Temporal Data
Sindhu Kishore and Hangfeng He
In NAACL 2024 (short papers). [pdf] -
Robust Learning Rate Selection for Stochastic Optimization via Splitting Diagnostic
Matteo Sordello, Niccolo Dalmasso, Hangfeng He, and Weijie Su
In TMLR 2024. [pdf]
2023
-
A Law of Data Separation in Deep Learning
Hangfeng He and Weijie Su
In PNAS 2023 (direct submission). [pdf] [code] -
On Regularization and Inference with Label Constraints
Kaifu Wang, Hangfeng He, Tin Nguyen, Piyush Kumar, and Dan Roth
In ICML 2023. [pdf] -
Incidental Supervision for Natural Language Understanding
Hangfeng He
PhD dissertation, University of Pennsylvania, 2023. [pdf]
2022
-
Transfer Learning via Representation Learning
Mohammad Rostami, Hangfeng He, Muhao Chen, and Dan Roth
In Federated and Transfer Learning 2022 (book chapter). [pdf] -
Weighted Training for Cross-Task Learning
Shuxiao Chen, Koby Crammer, Hangfeng He, Dan Roth, and Weijie Su (alphabetical order)
In ICLR 2022 (oral presentation). [pdf] [code]
2021
-
Exploring Deep Neural Networks via Layer-Peeled Model: Minority Collapse in Imbalanced Training
Cong Fang, Hangfeng He, Qi Long, and Weijie Su (alphabetical order)
In PNAS 2021 (direct submission). [pdf] [code] -
Foreseeing the Benefits of Incidental Supervision
Hangfeng He, Mingyuan Zhang, Qiang Ning, and Dan Roth
In EMNLP 2021. [pdf] [code] -
Toward Better Generalization Bounds with Locally Elastic Stability
Zhun Deng, Hangfeng He, and Weijie Su
In ICML 2021. [pdf]
2020
-
QANom: Question-Answer driven SRL for Nominalizations
Ayal Klein, Jonathan Mamou, Valentina Pyatkin, Daniela Brook Weiss, Hangfeng He, Dan Roth, Luke Zettlemoyer, and Ido Dagan
In COLING 2020. [pdf] -
Label-Aware Neural Tangent Kernel: Toward Better Generalization and Local Elasticity
Shuxiao Chen, Hangfeng He, and Weijie Su (alphabetical order)
In NeurIPS 2020. [pdf] [code] -
Towards Understanding the Dynamics of the First-Order Adversaries
Zhun Deng, Hangfeng He, Jiaoyang Huang, and Weijie Su
In ICML 2020. [pdf] -
QuASE: Question-Answer Driven Sentence Encoding
Hangfeng He, Qiang Ning, and Dan Roth
In ACL 2020. [pdf] [code] [demo] -
Understanding Spatial Relations through Multiple Modalities
Soham Dan, Hangfeng He, and Dan Roth
In LREC 2020 (short papers). [pdf] -
The Local Elasticity of Neural Networks
Hangfeng He and Weijie Su
In ICLR 2020. [pdf] [code]
2019
- Partial or Complete, That’s The Question
Qiang Ning, Hangfeng He, Chuchu Fan, and Dan Roth
In NAACL-HLT 2019. [pdf]
2018
- Cross-Domain and Semi-Supervised Named Entity Recognition in Chinese Social Media: A Unified Model
Jingjing Xu, Hangfeng He, Xu Sun, Xuancheng Ren, and Sujian Li
In TASLP 2018. [pdf]
2017
-
Detecting negation scope is easy, except when it isn’t
Federico Fancellu, Adam Lopez, Bonnie Webber, and Hangfeng He
In EACL 2017 (short papers). [pdf] -
F-Score Driven Max Margin Neural Network for Named Entity Recognition in Chinese Social Media
Hangfeng He and Xu Sun
In EACL 2017 (short papers). [pdf] -
A Unified Model for Cross-Domain and Semi-Supervised Named Entity Recognition in Chinese Social Media
Hangfeng He and Xu Sun
In AAAI 2017. [pdf] -
Neural Networks for Negation Cue Detection in Chinese
Hangfeng He, Federico Fancellu, and Bonnie Webber
In SemBEaR workshop 2017. [pdf]