Jan Buys
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Jan Buys has a total of 47 co-authors. Here are the top co-authors by number of shared publications:
Phil Blunsom
5 shared publicationsYejin Choi
School of Computer Science & Engineering Microsoft Research ♢ Allen Institute for Artificial Intelligence University of Washington ♣ Stanford University ♡ University of Cape Town
5 shared publicationsAri Holtzman
School of Computer Science & Engineering Microsoft Research ♢ Allen Institute for Artificial Intelligence University of Washington ♣ Stanford University ♡ University of Cape Town
4 shared publications- Francois Meyer
Department of Computer Science University of Cape Town
3 shared publications Edan Toledo
Department of Computer Science and Technology University of Cambridge
2 shared publications- Jonathan Phillip Shock
Department of Mathematics and Applied Mathematics University of Cape Town
2 shared publications
Latest Publications
- 25 November 2025
- Published by Springer Nature
- p. 144-158
Investigating the Impact of Multilingual Pre-trained Speech Models on Gender Bias in ASR for Low Resource African Languages
- 25 November 2025
- Published by Springer Nature
- p. 144-158
Investigating the Impact of Multilingual Pre-trained Speech Models on Gender Bias in ASR for Low Resource African Languages
- Open Access
- 1 January 2025
- Published by Association for Computational Linguistics (ACL)
- p. 483-495
Cross-Lingual Knowledge Augmentation for Mitigating Generic Overgeneralization in Multilingual Language Models
Sello Ralethe, Jan Buys. Proceedings of the 5th Workshop on Multilingual Representation Learning (MRL 2025). 2025.
- 1 January 2025
- Published by Association for Computational Linguistics (ACL)
- p. 483-495
Open AccessCross-Lingual Knowledge Augmentation for Mitigating Generic Overgeneralization in Multilingual Language Models
Sello Ralethe, Jan Buys. Proceedings of the 5th Workshop on Multilingual Representation Learning (MRL 2025). 2025.
- Open Access
A Systematic Analysis of Subwords and Cross-Lingual Transfer in Multilingual Translation
Multilingual modelling can improve machine translation for low-resource languages, partly through shared subword representations. This paper studies the role of subword segmentation in cross-lingual transfer. We systematically compare the efficacy of several subword methods in promoting synergy and preventing interference across different linguistic typologies. Our findings show that subword regularisation boosts synergy in multilingual modelling, whereas BPE more effectively facilitates transfer during cross-lingual fine-tuning. Notably, our results suggest that differences in orthographic word boundary conventions (the morphological granularity of written words) may impede cross-lingual transfer more significantly than linguistic unrelatedness. Our study confirms that decisions around subword modelling can be key to optimising the benefits of multilingual modelling.
Open AccessA Systematic Analysis of Subwords and Cross-Lingual Transfer in Multilingual Translation
Multilingual modelling can improve machine translation for low-resource languages, partly through shared subword representations. This paper studies the role of subword segmentation in cross-lingual transfer. We systematically compare the efficacy of several subword methods in promoting synergy and preventing interference across different linguistic typologies. Our findings show that subword regularisation boosts synergy in multilingual modelling, whereas BPE more effectively facilitates transfer during cross-lingual fine-tuning. Notably, our results suggest that differences in orthographic word boundary conventions (the morphological granularity of written words) may impede cross-lingual transfer more significantly than linguistic unrelatedness. Our study confirms that decisions around subword modelling can be key to optimising the benefits of multilingual modelling.
- Open Access
Triples-to-isiXhosa (T2X): Addressing the Challenges of Low-Resource Agglutinative Data-to-Text Generation
Most data-to-text datasets are for English, so the difficulties of modelling data-to-text for low-resource languages are largely unexplored. In this paper we tackle data-to-text for isiXhosa, which is low-resource and agglutinative. We introduce Triples-to-isiXhosa (T2X), a new dataset based on a subset of WebNLG, which presents a new linguistic context that shifts modelling demands to subword-driven techniques. We also develop an evaluation framework for T2X that measures how accurately generated text describes the data. This enables future users of T2X to go beyond surface-level metrics in evaluation. On the modelling side we explore two classes of methods - dedicated data-to-text models trained from scratch and pretrained language models (PLMs). We propose a new dedicated architecture aimed at agglutinative data-to-text, the Subword Segmental Pointer Generator (SSPG). It jointly learns to segment words and copy entities, and outperforms existing dedicated models for 2 agglutinative languages (isiXhosa and Finnish). We investigate pretrained solutions for T2X, which reveals that standard PLMs come up short. Fine-tuning machine translation models emerges as the best method overall. These findings underscore the distinct challenge presented by T2X: neither well-established data-to-text architectures nor customary pretrained methodologies prove optimal. We conclude with a qualitative analysis of generation errors and an ablation study.
Open AccessTriples-to-isiXhosa (T2X): Addressing the Challenges of Low-Resource Agglutinative Data-to-Text Generation
Most data-to-text datasets are for English, so the difficulties of modelling data-to-text for low-resource languages are largely unexplored. In this paper we tackle data-to-text for isiXhosa, which is low-resource and agglutinative. We introduce Triples-to-isiXhosa (T2X), a new dataset based on a subset of WebNLG, which presents a new linguistic context that shifts modelling demands to subword-driven techniques. We also develop an evaluation framework for T2X that measures how accurately generated text describes the data. This enables future users of T2X to go beyond surface-level metrics in evaluation. On the modelling side we explore two classes of methods - dedicated data-to-text models trained from scratch and pretrained language models (PLMs). We propose a new dedicated architecture aimed at agglutinative data-to-text, the Subword Segmental Pointer Generator (SSPG). It jointly learns to segment words and copy entities, and outperforms existing dedicated models for 2 agglutinative languages (isiXhosa and Finnish). We investigate pretrained solutions for T2X, which reveals that standard PLMs come up short. Fine-tuning machine translation models emerges as the best method overall. These findings underscore the distinct challenge presented by T2X: neither well-established data-to-text architectures nor customary pretrained methodologies prove optimal. We conclude with a qualitative analysis of generation errors and an ablation study.
- Open Access
- 1 January 2024
- Published by Association for Computational Linguistics (ACL)
- p. 12215-12229
Multipath parsing in the brain
Berta Franzluebbers, Donald Dunagan, Miloš Stanojević, Jan Buys, John Hale. Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2024.
- 1 January 2024
- Published by Association for Computational Linguistics (ACL)
- p. 12215-12229
Open AccessMultipath parsing in the brain
Berta Franzluebbers, Donald Dunagan, Miloš Stanojević, Jan Buys, John Hale. Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2024.
- Open Access
- 1 January 2024
- Published by Association for Computational Linguistics (ACL)
- p. 2194-2200
A Systematic Analysis of Subwords and Cross-Lingual Transfer in Multilingual Translation
Francois Meyer, Jan Buys. Findings of the Association for Computational Linguistics: NAACL 2024. 2024.
- 1 January 2024
- Published by Association for Computational Linguistics (ACL)
- p. 2194-2200
Open AccessA Systematic Analysis of Subwords and Cross-Lingual Transfer in Multilingual Translation
Francois Meyer, Jan Buys. Findings of the Association for Computational Linguistics: NAACL 2024. 2024.
- Open Access
Subword Segmental Machine Translation: Unifying Segmentation and Target Sentence Generation
Subword segmenters like BPE operate as a preprocessing step in neural machine translation and other (conditional) language models. They are applied to datasets before training, so translation or text generation quality relies on the quality of segmentations. We propose a departure from this paradigm, called subword segmental machine translation (SSMT). SSMT unifies subword segmentation and MT in a single trainable model. It learns to segment target sentence words while jointly learning to generate target sentences. To use SSMT during inference we propose dynamic decoding, a text generation algorithm that adapts segmentations as it generates translations. Experiments across 6 translation directions show that SSMT improves chrF scores for morphologically rich agglutinative languages. Gains are strongest in the very low-resource scenario. SSMT also learns subwords that are closer to morphemes compared to baselines and proves more robust on a test set constructed for evaluating morphological compositional generalisation.
Open AccessSubword Segmental Machine Translation: Unifying Segmentation and Target Sentence Generation
Subword segmenters like BPE operate as a preprocessing step in neural machine translation and other (conditional) language models. They are applied to datasets before training, so translation or text generation quality relies on the quality of segmentations. We propose a departure from this paradigm, called subword segmental machine translation (SSMT). SSMT unifies subword segmentation and MT in a single trainable model. It learns to segment target sentence words while jointly learning to generate target sentences. To use SSMT during inference we propose dynamic decoding, a text generation algorithm that adapts segmentations as it generates translations. Experiments across 6 translation directions show that SSMT improves chrF scores for morphologically rich agglutinative languages. Gains are strongest in the very low-resource scenario. SSMT also learns subwords that are closer to morphemes compared to baselines and proves more robust on a test set constructed for evaluating morphological compositional generalisation.
- Open Access
- 1 January 2023
- Published by Association for Computational Linguistics (ACL)
- p. 1230-1242
Policy-based Reinforcement Learning for Generalisation in Interactive Text-based Environments
Edan Toledo, Jan Buys, Jonathan Shock. Proceedings of the 17th Conference of the European Chapter of the Association for Computational Linguistics. 2023.
- 1 January 2023
- Published by Association for Computational Linguistics (ACL)
- p. 1230-1242
Open AccessPolicy-based Reinforcement Learning for Generalisation in Interactive Text-based Environments
Edan Toledo, Jan Buys, Jonathan Shock. Proceedings of the 17th Conference of the European Chapter of the Association for Computational Linguistics. 2023.
- 28 November 2022
- Published by Springer Nature
- p. 291-305
Self-Supervised Text Style Transfer with Rationale Prediction and Pretrained Transformers
- 28 November 2022
- Published by Springer Nature
- p. 291-305
Self-Supervised Text Style Transfer with Rationale Prediction and Pretrained Transformers
- 28 November 2022
- Published by Springer Nature
- p. 154-168
From GNNs to Sparse Transformers: Graph-Based Architectures for Multi-hop Question Answering
- 28 November 2022
- Published by Springer Nature
- p. 154-168
From GNNs to Sparse Transformers: Graph-Based Architectures for Multi-hop Question Answering