Papers by Joel Niklaus
Parity-Aware Byte-Pair Encoding: Improving Cross-lingual Fairness in Tokenization (2026.acl-long)
Copied to clipboard
Negar Foroutan, Clara Meister, Debjit Paul, Joel Niklaus, Sina Ahmadi, Antoine Bosselut, Rico Sennrich
| Challenge: | Tokenization is the first step of most NLP pipelines. |
| Approach: | They propose a parity-aware byte pair encoder that maximizes the compression gain of the currently worst-compressed language for cross-lingual parity. |
| Outcome: | a new algorithm reduces tokenization inequality by 89% compared to classical BPE . the proposed algorithm is based on a fair-max rule that maximizes the compression gain of the currently worst-compressed language . |
Unlocking Legal Knowledge: A Multilingual Dataset for Judicial Summarization in Switzerland (2025.findings-emnlp)
Copied to clipboard
Luca Rolshoven, Vishvaksenan Rasiah, Srinanda Brügger Bose, Sarah Hostettler, Lara Burkhalter, Matthias Stürmer, Joel Niklaus
| Challenge: | a dataset of 20K rulings from the Swiss Federal Supreme Court is lacking in legal headnotes due to the high cost of manual annotation. |
| Approach: | They propose a dataset that contains 20K rulings from the Swiss Federal Supreme Court . they fine-tune open models and compare them to larger general-purpose and reasoning-tunned LLMs . |
| Outcome: | The proposed dataset contains 20K rulings from the Swiss Federal Supreme Court with headnotes in German, French, and Italian. |
SwiLTra-Bench: The Swiss Legal Translation Benchmark (2025.acl-long)
Copied to clipboard
Joel Niklaus, Jakob Merane, Luka Nenadic, Sina Ahmadi, Yingqiang Gao, Cyrill A. H. Chevalley, Claude Humbel, Christophe Gösken, Lorenzo Tanzi, Thomas Lüthi, Stefan Palombo, Spencer Poff, Boling Yang, Nan Wu, Matthew Guillod, Robin Mamié, Daniel Brunner, Julio Pereyra, Niko Grupen
| Challenge: | In Switzerland legal translation relies on legal experts who must be both legal experts and skilled translators—creating bottlenecks and impacting effective access to justice. |
| Approach: | They propose a multilingual benchmarking system that evaluates Swiss legal translation systems based on 180K aligned Swiss legal translator pairs . they show frontier models achieve superior translation performance across all document types while specialized translation systems excel specifically in laws but under-perform in headnotes. |
| Outcome: | The proposed model outperforms specialized models in laws but underperform in headnotes. |
ConLoan: A Contrastive Multilingual Dataset for Evaluating Loanwords (2025.acl-long)
Copied to clipboard
Sina Ahmadi, Micha David Hess, Elena Álvarez-Mellado, Alessia Battisti, Cui Ding, Anne Göhring, Yingqiang Gao, Zifan Jiang, Andrianos Michail, Peshmerge Morad, Joel Niklaus, Maria Christina Panagiotopoulou, Stefano Perrella, Juri Opitz, Anastassia Shaitarova, Rico Sennrich
| Challenge: | Lexical borrowing is a ubiquitous linguistic phenomenon influenced by geopolitical, societal, and technological factors. |
| Approach: | They propose a novel contrastive dataset comprising sentences with and without loanwords across 10 languages to examine how machine translation and language models process loanword . |
| Outcome: | The proposed dataset shows that state-of-the-art models prefer loanwords over native terms and exhibit varying performance across languages. |
LEXTREME: A Multi-Lingual and Multi-Task Benchmark for the Legal Domain (2023.findings-emnlp)
Copied to clipboard
| Challenge: | Recent advances in legal NLP have led to a rapid growth of the field . however, many benchmarks are available only in English and no multilingual benchmark exists . |
| Approach: | They propose to use 11 datasets covering 24 languages to compare NLP models. |
| Outcome: | The proposed benchmarks show that even the best baseline only achieves modest results and ChatGPT struggles with many tasks. |
LegalLens: Leveraging LLMs for Legal Violation Identification in Unstructured Text (2024.eacl-long)
Copied to clipboard
Dor Bernsohn, Gil Semo, Yaron Vazana, Gila Hayat, Ben Hagag, Joel Niklaus, Rohit Saha, Kyryl Truskovskyi
| Challenge: | a recent study focused on detecting legal violations within unstructured textual data . a similar study focused only on associating violations with potentially affected individuals . |
| Approach: | They constructed two datasets using Large Language Models (LLMs) they publicize the results to advance legal natural language processing research . |
| Outcome: | The proposed datasets and the code used for the experiments have been released to advance legal natural language processing (NLP) |
Anonymity at Risk? Assessing Re-Identification Capabilities of Large Language Models in Court Decisions (2024.findings-naacl)
Copied to clipboard
| Challenge: | Despite high re-identification rates on Wikipedia, even the best LLMs struggled with court decisions. |
| Approach: | They construct an anonymized Wikipedia dataset to investigate re-identification risks . they also introduce new metrics to measure performance . |
| Outcome: | The proposed model can be used to identify individuals in court decisions, but it fails in the vast majority of cases. |
An Empirical Study on Cross-X Transfer for Legal Judgment Prediction (2022.aacl-main)
Copied to clipboard
| Challenge: | Cross-lingual transfer learning is understudied in legal NLP but not in legal Judgment Prediction (LJP). |
| Approach: | They explore cross-lingual transfer learning techniques on legal JP using a trilingual Swiss-Judgment-Prediction dataset and adapter-based fine-tuning. |
| Outcome: | The proposed methods improve the model’s performance by augmenting the training dataset with machine-translated versions of the original documents, using a 3 larger training corpus. |
From Citations to Criticality: Predicting Legal Decision Influence in the Multilingual Swiss Jurisprudence (2025.acl-short)
Copied to clipboard
| Challenge: | Existing approaches to evaluating the importance of legal cases are manual and resource-intensive. |
| Approach: | They propose a dataset that uses two-tier labels to evaluate case criticality . they use the LD-Label to identify cases published as Leading Decisions and the Citation-L Label to rank cases by their citation frequency and recency. |
| Outcome: | The Criticality Prediction dataset outperforms existing approaches to evaluate case criticality . the proposed model outperformed the existing models in a zero-shot setting . |
Resolving Legalese: A Multilingual Exploration of Negation Scope Resolution in Legal Documents (2024.lrec-main)
Copied to clipboard
| Challenge: | Negation scope resolution is a challenging task for NLP because of the complexity of legal texts and lack of annotated in-domain negation corpora. |
| Approach: | They propose to use annotated court decisions to improve negation scope resolution . they release annotations in german, french, and italian to train models without legal data . |
| Outcome: | The proposed models achieve token-level F1-scores of up to 86.7% in zero-shot and multilingual settings. |
Towards Explainability and Fairness in Swiss Judgement Prediction: Benchmarking on a Multilingual Dataset (2024.lrec-main)
Copied to clipboard
| Challenge: | Using Swiss Judgement Prediction, we evaluate the explainability of state-of-the-art monolingual and multilingual LJP models. |
| Approach: | They propose an occlusion-based approach to evaluate the explainability performance of legal judgement prediction models using Swiss Judgement Prediction, the only available multilingual LJP dataset. |
| Outcome: | The proposed framework allows us to quantify the influence of lower court information on model predictions, exposing current models’ biases. |
LawInstruct: A Resource for Studying Language Model Adaptation to the Legal Domain (2025.findings-naacl)
Copied to clipboard
Joel Niklaus, Lucia Zheng, Arya D. McCarthy, Christopher Hahn, Brian M Rosen, Peter Henderson, Daniel E. Ho, Garrett Honke, Percy Liang, Christopher D Manning
| Challenge: | In general, instruction tuning is important for direct user interaction, but the legal domain is underrepresented in typical instruction datasets. |
| Approach: | They aggregate 58 annotated legal datasets and write instructions for each to create LawInstruct. |
| Outcome: | The proposed model improves on LegalBench across all model sizes, but no drop in MMLU. |