Junwei Bao (鲍军威)

a photo

I am the leader of LLM training team at zuoyebang.com, focusing on ai-writing and ai-correcting related tasks which serve more than 10M users through apps including ‘作业帮’, ‘光速写作’ and ‘快对作业’. My team trains 1.5B, 8B, 12B, 13B, 70B, and 405B models using Pretraining, SFT, Preference Optimization, and Reject Sampling techniques. Previously, I was an NLP researcher at JD AI Research and an engineer at the Intelligent Service and Product department in JD Technology (JDT:京东科技). I have published more than 40 papers in top-tier conferences and journals, e.g., ICML, ACL, AAAI, EMNLP, and NAACL. My research topics include LLM, writing, dialog, question answering, reasoning, and machine reading comprehension. You can reach me through NAME at gmail dot com (NAME=baojunwei001).

Links

Education

I am from Anshan in Liaoning Province. I received my Ph.D. degree in Computer Science and Technology from Harbin Institute of Technology (HIT:哈尔滨工业大学) at January 03, 2019, after 6 years of NLP research in MSRA. I earned my Ph.D. degree under the instructions of Prof. Ming Zhou and Tiejun Zhao and mentored by Nan Duan. After that, I was a researcher at JD AI Research at JD.com, Beijing, working with Prof. Xiaodong He and Youzheng Wu.

News

05/2024: Our Interactive-KBQA was accepted by ACL 2024.
04/2023: Our SegCLIP was accepted by ICML 2023.
10/2022: Our Mars achieves the SOTA results (combined score 103.4) on MultiWoz leaderboard.
02/2022: Our UniRPG (EMNLP 2022) achieves the Top-1 on TAT-QA leaderboard.
02/2022: Our OPERA (NAACL 2022) achieves the Top-1 on DROP leaderboard.
12/2021: Our MuGER² (EMNLP 2022-Findings) achieves the Top-2 on HybridQA leaderboard.
05/2021: Our P³LM (EMNLP 2022-Findings) achieves the Top-2 on GLGE leaderboard.
01/2021: Our RoR (EMNLP 2021-Findings) achieves the Top-1 on QuAC leaderboard.

Opening Positions

I have some intern positions available at zuoyebang.com. If you have any interests, welcome to drop me an email.

Professional Activities

Area Chair (or SPC): IJCAI 2021.

Selected Publications

For full publication list, please check publication page.

[Paper] [CCF-A | 清华A] Guanming Xiong, Junwei Bao, Wen Zhao: Interactive-KBQA: Multi-Turn Interactions for Knowledge Base Question Answering with Large Language Models. ACL 2024
[Paper] [CCF-B | 清华A] Yongwei Zhou, Junwei Bao, Youzheng Wu, Xiaodong He, Tiejun Zhao: Operation-Augmented Numerical Reasoning for Question Answering. IEEE/ACM TASLP 2023
[Paper] [CCF-A | 清华A] Huaishao Luo, Junwei Bao*, Youzheng Wu, Xiaodong He, Tianrui Li: SegCLIP: Patch Aggregation with Learnable Centers for Open-Vocabulary Semantic Segmentation. ICML 2023
[Paper] [CCF-A | 清华A] Haoning Zhang, Junwei Bao, Haipeng Sun, Youzheng Wu, Wenye Li, Shuguang Cui, Xiaodong He: MoNET: Tackle State Momentum via Noise-Enhanced Training for Dialogue State Tracking. ACL(Findings) 2023
[Paper] [CCF-A | 清华A] Yu Lu, Junwei Bao, Zichen Ma, Xiaoguang Han, Youzheng Wu, Shuguang Cui, Xiaodong He: AUGUST: an Automatic Generation Understudy for Synthesizing Conversational Recommendation Datasets. ACL(Findings) 2023
[Paper] [CCF-A | 清华A] Haipeng Sun, Junwei Bao, Youzheng Wu, Xiaodong He: Mars: Modeling Context & State Representations with Contrastive Learning for End-to-End Task-Oriented Dialog. ACL(Findings) 2023
[Paper | Code] [CCF-B | 清华A] Junwei Bao*, Yifan Wang, Jiangyong Ying, Yeyun Gong, Jing Zhao, Youzheng Wu, Xiaodong He: P³LM: Probabilistically Permuted Prophet Language Modeling for Generative Pre-Training. EMNLP (Findings) 2022
[Paper | Code] [CCF-B | 清华A] Yongwei Zhou, Junwei Bao*, Chaoqun Duan, Youzheng Wu, Xiaodong He, Tiejun Zhao: UniRPG: Unified Discrete Reasoning over Table and Text as Program Generation. EMNLP 2022
[Paper | Code] [CCF-B | 清华A] Yingyao Wang, Junwei Bao*, Chaoqun Duan, Youzheng Wu, Xiaodong He, Tiejun Zhao: MuGER²: Multi-Granularity Evidence Retrieval and Reasoning for Hybrid Question Answering. EMNLP (Findings) 2022
[Paper | Code] [CCF-B | 清华B] Guanming Xiong, Junwei Bao*, Wen Zhao, Youzheng Wu, Xiaodong He: AutoQGS: Auto-Prompt for Low-Resource Knowledge-based Question Generation from SPARQL. CIKM 2022
[Paper | Code] [CCF-B | 清华B] Yongwei Zhou, Junwei Bao*, Chaoqun Duan, Haipeng Sun, Jiahui Liang, Yifan Wang, Jing Zhao, Youzheng Wu, Xiaodong He, Tiejun Zhao: OPERA: Operation-Pivoted Discrete Reasoning over Text. NAACL 2022
[Paper | Code] [CCF-B | 清华B] Haipeng Sun, Junwei Bao*, Youzheng Wu, Xiaodong He: BORT: Back and Denoising Reconstruction for End-to-End Task-Oriented Dialog. NAACL (Findings) 2022
[Paper | Code] [CCF-B | 清华B] Yifan Wang, Jing Zhao, Junwei Bao*, Chaoqun Duan, Youzheng Wu, Xiaodong He: LUNA: Learning Slot-Turn Alignment for Dialogue State Tracking. NAACL 2022
[Paper | Code] [CCF-B | 清华B] Guangyi Liu, Zichao Yang, Tianhua Tao, Xiaodan Liang, Junwei Bao, Xiaodong He, Shuguang Cui, Zhiting Hu: Don’t Take It Literally: An Edit-Invariant Sequence Loss for Text Generation. NAACL 2022
[Paper | Code] [CCF-A | 清华A] Jing Zhao, Yifan Wang, Junwei Bao*, Youzheng Wu, Xiaodong He: Fine- and Coarse-Granularity Hybrid Self-Attention for Efficient BERT. ACL 2022
[Paper | Code] [CCF-A | 清华A] Yu Lu, Junwei Bao*, Yan Song, Zichen Ma, Shuguang Cui, Youzheng Wu, Xiaodong He: RevCore: Review-Augmented Conversational Recommendation. ACL (Findings) 2021
[Paper | Code] [CCF-B | 清华A] Jing Zhao, Junwei Bao*, Yifan Wang, Yongwei Zhou, Youzheng Wu, Xiaodong He, Bowen Zhou: RoR: Read-over-Read for Long Document Machine Reading Comprehension. EMNLP (Findings) 2021
[Paper | Code] [CCF-B | 清华B] Jing Zhao, Junwei Bao*, Yifan Wang, Youzheng Wu, Xiaodong He, Bowen Zhou: SGG: Learning to Select, Guide, and Generate for Keyphrase Generation. NAACL 2021
[Paper | Code] [CCF-B | 清华B] Yingyao Wang, Junwei Bao*, Guangyi Liu, Youzheng Wu, Xiaodong He, Bowen Zhou, Tiejun Zhao: Learning to Decouple Relations: Few-Shot Relation Classification with Entity-Guided Attention and Confusion-Aware Training. COLING 2020
[Paper | Dataset] [CCF-A | 清华B] Bo Shao, Yeyun Gong, Junwei Bao, Jianshu Ji, Guihong Cao, Xiaola Lin, Nan Duan: Weakly Supervised Multi-task Learning for Semantic Parsing. IJCAI 2019
[Paper] [CCF-B | 清华A] Junwei Bao*, Duyu Tang, Nan Duan, Zhao Yan, Ming Zhou, Tiejun Zhao: Text Generation From Tables. IEEE/ACM TASLP 2019
[Paper] [CCF-B | 清华A] Junwei Bao*, Yeyun Gong, Nan Duan, Ming Zhou, Tiejun Zhao: Question Generation With Doubly Adversarial Nets. IEEE/ACM TASLP 2018
[Paper | Dataset] [CCF-A | 清华A] Junwei Bao*, Duyu Tang, Nan Duan, Zhao Yan, Yuanhua Lv, Ming Zhou, Tiejun Zhao: Table-to-Text: Describing Table Region With Natural Language. AAAI 2018
[Paper | Dataset] [CCF-B | 清华B] Junwei Bao*, Nan Duan, Zhao Yan, Ming Zhou, Tiejun Zhao: Constraint-Based Question Answering with Knowledge Graph. COLING 2016
[Paper] [CCF-A | 清华A] Zhao Yan, Nan Duan, Junwei Bao, Peng Chen, Ming Zhou, Zhoujun Li, Jianshe Zhou: DocChat: An Information Retrieval Approach for Chatbot Engines Using Unstructured Documents. ACL 2016
[Paper] [CCF-B | 清华B] Pengcheng Yin, Nan Duan, Ben Kao, Junwei Bao, Ming Zhou: Answering Questions with Complex Semantic Constraints on Open Knowledge Bases. CIKM 2015
[Paper] [CCF-A | 清华A] Junwei Bao*, Nan Duan, Ming Zhou, Tiejun Zhao: Knowledge-Based Question Answering as Machine Translation. ACL 2014

Collaborators

Interns/Students

My pleasure and great honor to work with these smart students.

Current Interns/Students

Past Interns/Students

Yingyao Wang (Ph.D. candidate at HIT, 08/2017 - 2023)
Yongwei Zhou (Ph.D. candidate at HIT, 08/2018 - 2023)
Guangyi Liu (Ph.D candidate at CUHK-Shenzhen, 08/2018 - 2023)
Haoning Zhang (Ph.D. candidate at CUHK-Shenzhen, 01/2019 - 2023)
Yu Lu (Ph.D. candidate at CUHK-Shenzhen, 01/2019 - 2023)
Jing Zhao (M.S. at Civil Aviation University, 1 ACL + 2 EMNLP Findings + 3 NAACL)
Shichang Ouyang (M.S. at Nothest University, 2 IJCAI)
Yatong Sun (M.S. at Nothest University)

Co-Authors

Xiaodong He, Youzheng Wu, Ming Zhou, Nan Duan, Tiejun Zhao, Bowen Zhou, Shuguang Cui, Duyu Tang, Yeyun Gong, and Zhao Yan.

Colleagues

Haipeng Sun, Chaoqun Duan, Huaishao Luo, Guanlin Li, and Jing Zhao