Qiwei Peng

Contact Me 🔎 - qipe@di.ku.dk

I am currently a Postdoc researcher at Coastal NLP Group at the University of Copenhagen, working with Prof. Anders Søggard. I obtained my PhD from the University of Sussex, supervised by Prof. David Weir and Dr. Julie Weeds (2019-2023). Before joining Sussex, I obtained my MSc in Natural Language Processing from the University of Edinburgh under supervision of Prof. Bonnie Webber and Dr. Thomas Kober (2017-18). I obtained BEng in Computer Science from Sichuan University (2013-2017).

Research interest: Crosslingual Generalization, Representation Learning, Trustworthy NLP

Apart from research, I enjoy reading novels and playing video games.

News

Sep 20, 2024	Four papers have been accepted to EMNLP 2024!!!
Feb 19, 2024	Our paper “HumanEval-XL: A Multilingual Code Generation Benchmark for Cross-Lingual Natural Language Generalization” has been accepted to LREC-COLING 2024!
May 16, 2023	Our paper “Testing Paraphrase Models on Recognising Sentence Pairs at Different Degrees of Semantic Overlap” has been accepted to the 12th *SEM conference (2023)!
Aug 17, 2022	Our Paper “Towards Structure-aware Paraphrase Identification with Phrase Alignment Using Sentence Encoders” has been accepted to COLING 2022!
Feb 24, 2022	Our paper “Predicate-Argument Based Bi-Encoder for Paraphrase Identification “ has been accepted to ACL 2022!

Selected Publications

EMNLP

Concept Space Alignment in Multilingual LLMs

Qiwei Peng, and Anders Søgaard

2024
LREC-COLING

HumanEval-XL: A Multilingual Code Generation Benchmark for Cross-Lingual Natural Language Generalization

Qiwei Peng*, Yekun Chai*, and Xuhong Li

In The 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation May 2024

URL
COLINGOral

Towards Structure-aware Paraphrase Identification with Phrase Alignment Using Sentence Encoders

Qiwei Peng, David Weir, and Julie Weeds

In Proceedings of the 29th International Conference on Computational Linguistics Oct 2022
ACLOral

Predicate-Argument Based Bi-Encoder for Paraphrase Identification

Qiwei Peng, David Weir, Julie Weeds, and Yekun Chai

In Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) May 2022

Abs URL

Paraphrase identification involves identifying whether a pair of sentences express the same or similar meanings. While cross-encoders have achieved high performances across several benchmarks, bi-encoders such as SBERT have been widely applied to sentence pair tasks. They exhibit substantially lower computation complexity and are better suited to symmetric tasks. In this work, we adopt a bi-encoder approach to the paraphrase identification task, and investigate the impact of explicitly incorporating predicate-argument information into SBERT through weighted aggregation. Experiments on six paraphrase identification datasets demonstrate that, with a minimal increase in parameters, the proposed model is able to outperform SBERT/SRoBERTa significantly. Further, ablation studies reveal that the predicate-argument based component plays a significant role in the performance gain.