News

  • Dec 1, 2021, the following paper: KGR4: Retrieval, Retrospect, Refine and Rethink for Commonsense Generation was accepted to the AAAI 2022.

  • Dec 1, 2021, the following paper: Frequency-Aware Contrastive Learning for Neural Machine Translation was accepted to the AAAI 2022.

  • Nov 11, 2021, our model RoBLEURT achieves a total of 5 Winners and 2 Second place results in all 8 XX-to-EN segment-level MT evalutation tasks of WMT2021.

  • Sep 9, 2021, the following paper: RoBLEURT Submission for WMT2021 Metrics Task was accepted to WMT 2021.

  • May 18, 2021, our model KGR^4 achieves 1st place in CommonGen leaderboard.

  • May 8, 2021, the following paper: BANG: Bridging Autoregressive and Non-autoregressive Generation with Large Scale Pretraining was accepted to the ICML 2021.

  • May 6, 2021, the following paper: Towards User-Driven Neural Machine Translation was accepted to the ACL 2021.

  • May 6, 2021, the following paper: POS-constrained Parallel Decoding for Non-autoregressive Generation was accepted to the ACL 2021.

  • May 6, 2021, the following paper: Bridging Subword Gaps in Pretrain-Finetune Paradigm for Natural Language Generation was accepted to the ACL 2021.

  • May 6, 2021, the following paper: GLGE: A New General Language Generation Evaluation Benchmark was accepted to the findings of ACL 2021.

  • Apr 10, 2021, the following paper: AnchiBERT: A Pre-Trained Model for Ancient Chinese Language Understanding and Generation was accepted to the IJCNN 2021.

  • Mar 31, 2021, the following paper: Mask Attention Networks: Rethinking and Strengthen Transformer was accepted to the NAACL 2021.

  • Mar 3, 2021, the following paper: ProphetNet: Predicting Future N-gram for Sequence-to-Sequence Pre-training was selected as Most Influential EMNLP Papers by paperdigest.org.

  • Nov 30, 2020, I was invited to be IJCAI’21 SPC.

  • Nov 14, 2020, we propose a new General Langugae Generation Evaluation benchmark called GLGE, the code and dataset available at here, and the leaderboard is available at here. Have a try! :)

  • Nov 6, 2020, our sequence-to-sequence pre-training model ProphetNet has been added into the Huggingface Transformers. Have a try! :)

  • Nov 5, 2020, the following paper: Herb-Know: Knowledge Enhanced Prescription Generation for Traditional Chinese Medicine was accepted to the BIBM-ITCM 2020.

  • Sep 15, 2020, the following paper: Tell Me How to Ask Again: Question Data Augmentation with Controllable Rewriting in Continuous Space was accepted to the EMNLP 2020.

  • Sep 15, 2020, the following paper: Diverse, Controllable, and Keyphrase-Aware: A Corpus and Method for News Multi-Headline Generation was accepted to the EMNLP 2020.

  • Sep 15, 2020, the following paper: ProphetNet: Predicting Future N-gram for Sequence-to-Sequence Pre-training was accepted to the findings of EMNLP 2020.

  • Sep 1, 2020, the following paper: Exploration on the Generation of Chinese Palindrome Poetry was accepted to the ICONIP 2020.

  • Apr 4, 2020, the following paper: RikiNet: Reading Wikipedia Pages for Natural Question Answering was accepted to the ACL 2020.

  • Mar 21, 2020, the following paper: Generating Chinese Poetry from Images via Concrete and Abstract Information was accepted to the IJCNN 2020.

  • Nov 29, 2019, Our model RikiNet achieves 1st place in Natural Question (NQ) leaderboard.

  • Nov 20, 2019, the demo video of the following paper: Deep Poetry: A Chinese Classical Poetry Generation System (AAAI 2020 demo) is available at here.

  • Nov 16, 2019, the code and data of the following paper: Revision in Continuous Space: Unsupervised Text Style Transfer without Adversarial Learning (AAAI 2020) are available at here.

  • Nov 11, 2019, the following paper: Revision in Continuous Space: Unsupervised Text Style Transfer without Adversarial Learning was accepted to the AAAI 2020.

  • Nov 04, 2019, I got the “Academic Star” Award of Sichuan University

  • Oct 25, 2019, the following paper: Deep Poetry: A Chinese Classical Poetry Generation System was accepted to the AAAI 2020 Demonstrations Program.

  • Oct 10, 2019, I got the National Scholarship for Ph.D. Students.

  • Sep 10, 2019, the following paper: BFGAN: Backward and Forward Generative Adversarial Networks for Lexically Constrained Sentence Generation was accepted to the IEEE/ACM Transactions on Audio, Speech, and Language Processing (TASLP 2019).

  • June 24, 2019, I started my internship at NLC group of Microsoft Research Lab – Asia (MSRA).

  • June 20, 2019, the dataset of the following paper: Ancient-Modern Chinese Translation with a New Large Training Dataset (TALLIP 2019) are available at here.

  • May 26, 2019, the code and data of the following paper: TIGS: An Inference Algorithm for Text Infilling with Gradient Search (ACL 2019) are available at here.

  • May 24, 2019, the code and data of the following paper: Mu-Forcing: Training Variational Recurrent Autoencoders for Text Generation (TALLIP 2019) are available at here.

  • May 14, 2019, the following paper: TIGS: An Inference Algorithm for Text Infilling with Gradient Search was accepted to the ACL 2019.

  • May 06, 2019, the following paper: Mu-Forcing: Training Variational Recurrent Autoencoders for Text Generation was accepted to the ACM Transactions on Asian and Low-Resource Language Information Processing (TALLIP 2019).

  • Apr 15, 2019, the following paper: Ancient-Modern Chinese Translation with a New Large Training Dataset was accepted to the ACM Transactions on Asian and Low-Resource Language Information Processing (TALLIP 2019).