Publications

2025

1. EMNLP

   Understanding Subword Compositionality of Large Language Models

   Qiwei Peng, Yekun Chai, and Anders Søgaard

   In Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing Nov 2025
2. EMNLP

   Debiasing Multilingual LLMs in Cross-lingual Latent Space

   Qiwei Peng, Guimin Hu, Yekun Chai, and Anders Søgaard

   In Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing Nov 2025
3. EMNLPSystem Demo Track

   o-MEGA: Optimized Methods for Explanation Generation and Analysis

   Ľuboš Kriš, Jaroslav Kopčan, Qiwei Peng, Andrej Ridzik, Marcel Veselý, and Martin Tamajka

   In Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing: System Demonstrations Nov 2025
4. SemEval

   SemEval-2025 Task 7: Multilingual and Crosslingual Fact-Checked Claim Retrieval

   Qiwei Peng, Robert Moro, Michal Gregor, Ivan Srba, Simon Ostermann, Marian Simko, and 4 more authors

   In Proceedings of the 19th International Workshop on Semantic Evaluation (SemEval-2025) Aug 2025

2024

1. EMNLP

   Concept Space Alignment in Multilingual LLMs

   Qiwei Peng, and Anders Søgaard

   In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing Nov 2024
2. EMNLP

   FoodieQA: A Multimodal Dataset for Fine-Grained Understanding of Chinese Food Culture

   Wenyan Li, Xinyu Zhang, Jiaang Li, Qiwei Peng, Raphael Tang, Li Zhou, and 5 more authors

   In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing Nov 2024
3. EMNLP

   On Training Data Influence of GPT Models

   Qingyi Liu, Yekun Chai, Shuohuan Wang, Yu Sun, Qiwei Peng, and Hua Wu

   In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing Nov 2024
4. EMNLPFindings

   Tokenization Falling Short: On Subword Robustness in Large Language Models

   Yekun Chai, Yewei Fang, Qiwei Peng, and Xuhong Li

   In Findings of the Association for Computational Linguistics: EMNLP 2024 Nov 2024
5. 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

2023

1. *SEM

   Testing Paraphrase Models on Recognising Sentence Pairs at Different Degrees of Semantic Overlap

   Qiwei Peng, David Weir, and Julie Weeds

   In Proceedings of the 12th Joint Conference on Lexical and Computational Semantics (*SEM 2023) Jul 2023

   URL

2022

1. 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
2. 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.

2021

1. RepL4NLP@ACL

   Structure-aware Sentence Encoder in Bert-Based Siamese Network

   Qiwei Peng, David Weir, and Julie Weeds

   In Proceedings of the 6th Workshop on Representation Learning for NLP (RepL4NLP-2021) Aug 2021

   Abs URL

   Recently, impressive performance on various natural language understanding tasks has been achieved by explicitly incorporating syntax and semantic information into pre-trained models, such as BERT and RoBERTa. However, this approach depends on problem-specific fine-tuning, and as widely noted, BERT-like models exhibit weak performance, and are inefficient, when applied to unsupervised similarity comparison tasks. Sentence-BERT (SBERT) has been proposed as a general-purpose sentence embedding method, suited to both similarity comparison and downstream tasks. In this work, we show that by incorporating structural information into SBERT, the resulting model outperforms SBERT and previous general sentence encoders on unsupervised semantic textual similarity (STS) datasets and transfer classification tasks.

2. ACLFindings

   Representing Syntax and Composition with Geometric Transformations

   Lorenzo Bertolini, Julie Weeds, David Weir, and Qiwei Peng

   In Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021 Aug 2021

   URL
3. Behav Sci

   Aroused and Impulsive Effects of Colour Stimuli on Lateral and Logical Abilities

   Guobin Xia, Muzi Li, Philip Henry, Stephen Westland, Francisco Queiroz, Qiwei Peng, and 1 more author

   Behavioral Sciences Aug 2021