Biography

Xuancheng Ren is a graduate student in the School of Electronics Engineering and Computer Science at Peking University, where he is advised by Professor Xu Sun. He received the degree of Bachelor of Science in Computer Science from Peking University in 2017.

His research interests include natural language processing (NLP) and machine learning for NLP. Currently, he works on grounded language processing, which tries to incorporate various sources of knowledge (e.g., vision, commonsense, and semantics) into intelligent systems for reliable and explainable language processing. Previously, he had focused on techniques for efficient deep learning.

Interests

  • Natural Language Processing
    • Language Generation
    • NLP for Chinese
  • Machine Learning for NLP

Education

  • PhD Candidate in Computer Software and Theory

    Peking University

  • Bachelor of Science in Computer Science, 2017

    Peking University

Publications

A Global Past-Future Early Exit Method for Accelerating Inference of Pre-trained Language Models

Early exit mechanism aims to accelerate inference speed for large-scale pre-trained language models. The essential idea is exiting …
NAACL 2021 (to appear)
Code

Be Careful about Poisoned Word Embeddings: Exploring the Vulnerability of the Embedding Layers in NLP Models

Recent studies have revealed a security threat to natural language processing (NLP) models, called the Backdoor Attack. Victim models …
NAACL 2021 (to appear)
arXiv Code

Neural Network Surgery: Injecting Data Patterns into Pre-trained Models with Minimal Instance-wise Side Effects

Side effects during neural network tuning are typically measured by overall accuracy changes. However, we find that even with similar …
NAACL 2021 (to appear)

Collaborative Group Learning

Collaborative learning has successfully applied knowledge transfer to guiding a pool of small student networks towards robust local …
AAAI 2021 (to appear)
arXiv

Exploring the Vulnerability of Deep Neural Networks: A Study of Parameter Corruption

We argue that the vulnerability of model parameters is of crucial value to the study of model robustness and generalization but little …
AAAI 2021 (to appear)
URL arXiv

Multi-View Feature Representation for Dialogue Generation with Bidirectional Distillation

Neural dialogue models suffer from low-quality responses when interacted in practice, demonstrating difficulty in generalization over …
AAAI 2021 (to appear)
arXiv

Towards Semantics-Enhanced Pre-Training: Can Lexicon Definitions Help Learning Sentence Meanings?

Self-supervised pre-training techniques, relying on large amounts of text, have enabled rapid growth in bi-directional language …
AAAI 2021 (to appear)
URL Code

Prophet Attention: Predicting Attention with Future Attention for Improved Image Captioning

Recently, attention based models have been used extensively in image captioning and are expected to ground correct image regions with …
NeurIPS 2020 (to appear)
URL

Regularizing Dialogue Generation by Imitating Implicit Scenarios

Human dialogues are scenario-based and appropriate responses generally relate to the latent context knowledge entailed by the specific …
EMNLP 2020
URL DOI arXiv

Aligning Visual Regions and Textual Concepts for Semantic-Grounded Image Representations

In vision-and-language grounding problems, fine-grained representations of the image are considered to be of paramount importance. Most …
NeurIPS 2019
URL arXiv Code
Aligning Visual Regions and Textual Concepts for Semantic-Grounded Image Representations
See all publications