I am a researcher with over 10 years of well-balanced experience in academia and industry. I am now with Inf, a startup company working on large language model solutions for industry applications. I received my Ph.D. degree from Peking University, under the supervision of Prof. Ying Tan. Before that, I obtained my BS. from South China University of Technology. My current research mainly focuses on logical reasoning, especially neuro-symbolic methods. Previously, I mainly worked on learning latent variable models for NLP problems. We have some research intern positions available in Inf. If you are interested in LLM and logical reasoning, please contact me: wdxu@inftech.ai.

徐威迪 / wead_hsu@163.com / Scholar / Twitter / PubMed

News

• One paper is accepted by ICLR 2024 (first author).
• One paper is accepted by KDD 2023.
• I won the second prize of Beijing Science and Technology Award.

Blogs

• Introducing Neuro-symbolic Methods
• Efficient Softmax
• Backpropogate through Discrete Variables

Publications

[ICLR 2024] LogicMP: A Neuro-Symbolic Approach for Encoding First-order Logic Constraints.

• Weidi Xu, Jingwei Wang, Lele Xie, Jianshan He, Hongting Zhou, Taifeng Wang, Xiaopei Wan, Jingdong Chen, Chao Qu, Wei Chu.

• Slides / Poster / Code

[KDD 2023] Learning to Discover Various Simpson’s Paradoxes.

• Jingwei Wang, Jianshan He, Weidi Xu, Ruopeng Li, Wei Chu.

[EMNLP 2022] Extracting Trigger-sharing Events via an Event Matrix.

• Jun Xu, Weidi Xu, Mengshu Sun, Taifeng Wang, Wei Chu.

[WSDM 2020] Modeling Across-Context Attention For Long-Tail Query Classification in E-commerce.

• Junhao Zhang, Weidi Xu, Jianhui Ji, Xi Chen, Hongbo Deng, Keping Yang.

[EMNLP 2020] Question Directed Graph Attention Network for Numerical Reasoning over Text.

• Kunlong Chen, Weidi Xu, Xingyi Cheng, Zou Xiaochuan, Yuyu Zhang, Le Song†, Taifeng Wang, Yuan Qi, Wei Chu.
• SOTA results on DROP, outperforming GPT-4.

[SIGIR 2020] Symmetric Regularization based BERT for Pair-wise Semantic Reasoning.

• Weidi Xu, Xingyi Cheng, Kunlong Chen, Wei Wang, Bin Bi, Ming Yan, Chen Wu, Luo Si, Wei Chu, Taifeng Wang.

[ACL 2020] SpellGCN: Incorporating Phonological and Visual Similarities into Language Models for Chinese Spelling Check.

• Weidi Xu, Xingyi Cheng, Kunlong Chen, Shaohua Jiang, Feng Wang, Taifeng Wang, Wei Chu, Yuan Qi.

[CVPR 2020] Data-Efficient Semi-Supervised Learning by Reliable Edge Mining.

• Peibin Chen, Tao Ma, Xu Qin, Weidi Xu, Shuchang Zhou.

[TNNLS 2019] Semisupervised Text Classification by Variational Autoencoder.

• Weidi Xu, Haoze Sun, Chao Deng, Ying Tan.

[Neurocomputing 2019] Semi-supervised Target-level Sentiment Analysis via Variational Autoencoder.

• Weidi Xu, Ying Tan.

[CoNLL 2019] Variational Semi-supervised Aspect-term Sentiment Analysis via Transformer.

• Xingyi Cheng, Weidi Xu, Taifeng Wang, Wei Chu.

[CEC 2018] TextDream: Conditional Text Generation by Searching in the Semantic Space.

• Weidi Xu, Haoze Sun, Chao Deng, Ying Tan.

[AAAI 2017] Variational Autoencoders for Semi-supervised Text Classification.

• Weidi Xu, Haoze Sun, Chao Deng, Ying Tan.

[IJCNN 2016] Multi-digit Image Synthesis using Recurrent Conditional Variational Autoencoder.

• Haoze Sun, Weidi Xu, Chao Deng, Ying Tan.

[中国科学 2014] 基于平均增益模型的连续型 (1+1) 进化算法计算时间复杂性分析.

• 黄翰, 徐威迪, 张宇山, 林智勇, 郝志峰.