Tian-Xing He

Most-cited papers

1. Why Gradient Clipping Accelerates Training: A Theoretical Justification for Adaptivity

   J. Zhang, Tian-Xing He, Suvrit Sra, A. Jadbabaie

   2019International Conference on Learning Representations672 citations

   Semantic Scholar

   Save

2. Can Language Models Solve Graph Problems in Natural Language?

   Heng Wang, Shangbin Feng, Tian-Xing He, Zhaoxuan Tan +2 more

   2023Neural Information Processing Systems425 citations Open access

   Semantic Scholar

   Save

3. Reshaping deep neural network for fast decoding by node-pruning

   Tian-Xing He, Yuchen Fan, Y. Qian, Tian Tan +1 more

   2014IEEE International Conference on Acoustics, Speech, and Signal Processing121 citations

   Semantic Scholar

   Save

4. Negative Training for Neural Dialogue Response Generation

   Tian-Xing He, James R. Glass

   2019Annual Meeting of the Association for Computational Linguistics80 citations Open access

   Semantic Scholar

   Save

5. On the Blind Spots of Model-Based Evaluation Metrics for Text Generation

   Tian-Xing He, Jingyu (Jack) Zhang, Tianle Wang, Sachin Kumar +3 more

   2022Annual Meeting of the Association for Computational Linguistics75 citations Open access

   Semantic Scholar

   Save

6. Analyzing the Forgetting Problem in Pretrain-Finetuning of Open-domain Dialogue Response Models

   Tian-Xing He, Jun Liu, Kyunghyun Cho, Myle Ott +3 more

   2021Conference of the European Chapter of the Association for Computational Linguistics62 citations Open access

   Semantic Scholar

   Save

7. Knowledge Card: Filling LLMs' Knowledge Gaps with Plug-in Specialized Language Models

   Shangbin Feng, Weijia Shi, Yuyang Bai, Vidhisha Balachandran +2 more

   2023International Conference on Learning Representations60 citations

   Semantic Scholar

   Save

8. Exposure Bias versus Self-Recovery: Are Distortions Really Incremental for Autoregressive Text Generation?

   Tian-Xing He, Jingzhao Zhang, Zhiming Zhou, James R. Glass

   2019Conference on Empirical Methods in Natural Language Processing56 citations Open access

   Semantic Scholar

   Save

9. A comprehensive empirical study on bug characteristics of deep learning frameworks

   Yilin Yang, Tian-Xing He, Zhilong Xia, Yang Feng

   2022Information and Software Technology43 citations

   Semantic Scholar

   Save

10. A Systematic Characterization of Sampling Algorithms for Open-ended Language Generation

    Moin Nadeem, Tian-Xing He, Kyunghyun Cho, James R. Glass

    2020AACL42 citations Open access

    Semantic Scholar

    Save

11. Quantifying Exposure Bias for Neural Language Generation

    Tian-Xing He, J. Zhang, Zhi-Ming Zhou, James R. Glass

    2019arXiv.org37 citations

    Semantic Scholar

    Save

12. Exploiting LSTM structure in deep neural networks for speech recognition

    Tian-Xing He, J. Droppo

    2016IEEE International Conference on Acoustics, Speech, and Signal Processing37 citations

    Semantic Scholar

    Save

13. An Empirical Study of Transformer-Based Neural Language Model Adaptation

    Ke Li, Zhe Liu, Tian-Xing He, Hongzhao Huang +3 more

    2020IEEE International Conference on Acoustics, Speech, and Signal Processing32 citations

    Semantic Scholar

    Save

14. From Data Quality to Model Quality: An Exploratory Study on Deep Learning

    Tian-Xing He, Sheng-Cheng Yu, Zi-Yuan Wang, Jie-Qiong Li +1 more

    2019Asia-Pacific Symposium on Internetware32 citations

    Semantic Scholar

    Save

15. Joint Energy-based Model Training for Better Calibrated Natural Language Understanding Models

    Tian-Xing He, Bryan McCann, Caiming Xiong, Ehsan Hosseini-Asl

    2021Conference of the European Chapter of the Association for Computational Linguistics24 citations Open access

    Semantic Scholar

    Save

16. Detecting egregious responses in neural sequence-to-sequence models

    Tian-Xing He, James R. Glass

    2018International Conference on Learning Representations23 citations

    Semantic Scholar

    Save

17. Mining Python fix patterns via analyzing fine-grained source code changes

    Yilin Yang, Tian-Xing He, Yang Feng, Shaoying Liu +1 more

    2022Empirical Software Engineering19 citations

    Semantic Scholar

    Save

18. On training bi-directional neural network language model with noise contrastive estimation

    Tian-Xing He, Yu Zhang, J. Droppo, Kai Yu

    2016International Symposium on Chinese Spoken Language Processing18 citations Open access

    Semantic Scholar

    Save

19. Analyzing the Forgetting Problem in the Pretrain-Finetuning of Dialogue Response Models

    Tian-Xing He, Jun Liu, Kyunghyun Cho, Myle Ott +3 more

    201916 citations

    Semantic Scholar

    Save

20. An Empirical Study on Few-shot Knowledge Probing for Pretrained Language Models

    Tian-Xing He, Kyunghyun Cho, James R. Glass

    2021arXiv.org13 citations

    Semantic Scholar

    Save

21. AutoKG: Constructing Virtual Knowledge Graphs from Unstructured Documents for Question Answering

    S. Yu, Tian-Xing He, James R. Glass

    202112 citations

    Semantic Scholar

    Save

22. Revisiting Latent-Space Interpolation via a Quantitative Evaluation Framework

    Lu Mi, Tian-Xing He, C. Park, Hao Wang +2 more

    2021arXiv.org12 citations

    Semantic Scholar

    Save

23. Analysis of Gradient Clipping and Adaptive Scaling with a Relaxed Smoothness Condition

    J. Zhang, Tian-Xing He, Suvrit Sra, A. Jadbabaie

    2019arXiv.org10 citations

    Semantic Scholar

    Save

24. Mix-review: Alleviate Forgetting in the Pretrain-Finetune Framework for Neural Language Generation Models

    Tian-Xing He, Jun Liu, Kyunghyun Cho, Myle Ott +3 more

    2019arXiv.org9 citations

    Semantic Scholar

    Save

25. An investigation on DNN-derived bottleneck features for GMM-HMM based robust speech recognition

    Yongbin You, Y. Qian, Tian-Xing He, Kai Yu

    2015China Summit and International Conference on Signal and Information Processing8 citations

    Semantic Scholar

    Save

26. Recurrent neural network language model with structured word embeddings for speech recognition

    Tian-Xing He, Xu Xiang, Y. Qian, Kai Yu

    2015IEEE International Conference on Acoustics, Speech, and Signal Processing6 citations

    Semantic Scholar

    Save

27. Automatic model redundancy reduction for fast back-propagation for deep neural networks in speech recognition

    Y. Qian, Tian-Xing He, Wei Deng, Kai Yu

    2015IEEE International Joint Conference on Neural Network6 citations

    Semantic Scholar

    Save

28. Controlling the Focus of Pretrained Language Generation Models

    Jiabao Ji, Yoon Kim, James R. Glass, Tian-Xing He

    2022Findings5 citations Open access

    Semantic Scholar

    Save

29. PCFG-Based Natural Language Interface Improves Generalization for Controlled Text Generation

    Jingyu (Jack) Zhang, James R. Glass, Tian-Xing He

    2022STARSEM5 citations Open access

    Semantic Scholar

    Save

30. Constructing a Knowledge Graph from Unstructured Documents without External Alignment

    Seunghak Yu, Tian-Xing He, James R. Glass

    2020arXiv.org5 citations

    Semantic Scholar

    Save

31. Paragraph vector based topic model for language model adaptation

    Wen-Gong Jin, Kai Yu, Tian-Xing He, Y. Qian

    2015Interspeech5 citations

    Semantic Scholar

    Save

32. Multi-view LSTM Language Model with Word-Synchronized Auxiliary Feature for LVCSR

    Yue Wu, Tian-Xing He, Zhe-Huai Chen, Y. Qian +1 more

    2017China National Conference on Chinese Computational Linguistics4 citations

    Semantic Scholar

    Save

33. Q UANTIFYING E XPOSURE B IAS FOR O PEN - ENDED L ANGUAGE G ENERATION

    Tian-Xing He, Jingzhao Zhang, Zhi-Ming Zhou

    20203 citations

    Semantic Scholar

    Save