Matthias Grabmair

Sebastian Nagl, Matthias Grabmair

Evaluating large language models (LLMs) for legal reasoning requires workflows that span task design, expert annotation, model execution, and metric-based evaluation. In practice, these steps are split across platforms and scripts, limiting transparency, reproducibility, and participation by non-technical legal experts. We present the BenGER (Benchmark for German Law) framework, an open-source web platform that integrates task creation, collaborative annotation, configurable LLM runs, and evaluation with lexical, semantic, factual, and judge-based metrics. BenGER supports multi-organization projects with tenant isolation and role-based access control, and can optionally provide formative, reference-grounded feedback to annotators. We will demonstrate a live deployment showing end-to-end benchmark creation and analysis.

Sebastian Nagl, Ann-Kristin Mayrhofer, Martin Heidebach et al.

We introduce the BenGER (Benchmark for German Law) dataset for evaluating LLM systems on subsumption-based legal reasoning in German law. The BenGER dataset consists of three components: 596 exam-style free-text legal case tasks across multiple levels of legal education and 531 short doctrinal reasoning tasks. We evaluate 12 contemporary LLM systems -- closed flagship, efficiency-oriented, and open-weight -- across automatic and judge-based metrics. On a controlled validation subset of timed human-written solutions under both unaided and human--AI co-creation conditions, we contextualise model performance against these human baselines. We introduce a rubric-aligned LLM-as-a-Judge framework cross-validated against a multi-rater human-grading protocol (three blind reviews plus one author-informed creator review per solution). Our results show that replacing a blind human reviewer with the LLM judge degrades agreement with the full human pool no more than removing that reviewer altogether (Calderon r=0.96 vs.~r=0.96, matched n=30), that closed-flagship systems lead the leaderboard across all corpora, and that human--AI co-creation substantially outperforms unaided human work.

T. Y. S. S Santosh, Shanshan Xu, Oana Ichim et al.

This work demonstrates that Legal Judgement Prediction systems without expert-informed adjustments can be vulnerable to shallow, distracting surface signals that arise from corpus construction, case distribution, and confounding factors. To mitigate this, we use domain expertise to strategically identify statistically predictive but legally irrelevant information. We adopt adversarial training to prevent the system from relying on it. We evaluate our deconfounded models by employing interpretability techniques and comparing to expert annotations. Quantitative experiments and qualitative analysis show that our deconfounded model consistently aligns better with expert rationales than baselines trained for prediction only. We further contribute a set of reference expert annotations to the validation and testing partitions of an existing benchmark dataset of European Court of Human Rights cases.

T. Y. S. S Santosh, Oana Ichim, Matthias Grabmair

In this paper, we cast Legal Judgment Prediction on European Court of Human Rights cases into an article-aware classification task, where the case outcome is classified from a combined input of case facts and convention articles. This configuration facilitates the model learning some legal reasoning ability in mapping article text to specific case fact text. It also provides an opportunity to evaluate the model's ability to generalize to zero-shot settings when asked to classify the case outcome with respect to articles not seen during training. We devise zero-shot experiments and apply domain adaptation methods based on domain discrimination and Wasserstein distance. Our results demonstrate that the article-aware architecture outperforms straightforward fact classification. We also find that domain adaptation methods improve zero-shot transfer performance, with article relatedness and encoder pre-training influencing the effect.

T. Y. S. S Santosh, Marcel Perez San Blas, Phillip Kemper et al.

We report on an experiment in case outcome classification on European Court of Human Rights cases where our model first learns to identify the convention articles allegedly violated by the state from case facts descriptions, and subsequently uses that information to classify whether the court finds a violation of those articles. We assess the dependency between these two tasks at the feature and outcome level. Furthermore, we leverage a hierarchical contrastive loss to pull together article-specific representations of cases at the higher level, leading to distinctive article clusters. The cases in each article cluster are further pulled closer based on their outcome, leading to sub-clusters of cases with similar outcomes. Our experiment results demonstrate that, given a static pre-trained encoder, our models produce a small but consistent improvement in classification performance over single-task and joint models without contrastive loss.

Julius Vernie, Matthias Grabmair

Formalizing legal provisions promises machine-accessible law and automated legal reasoning, and recent LLMs make it tempting to generate such formalizations directly from statutory text. However, any formalization makes implicit interpretive choices whose consequences are hard to anticipate, especially if an LLM is the author. We present a method for systematically comparing different formalizations of the same legal provision by their inferences on individual cases. Given multiple formalizations of a provision, we match them at the node level, derive a shared interface for each pair from the matching, and use a SAT solver to enumerate the edge cases on which any two formalizations disagree. Selected edge cases are then verbalized into concrete factual scenarios that a legal expert can examine and act on. We apply our method to formalizations of ten EU provisions generated by nine frontier LLMs. We find that behavioral divergence between formalizations is essentially uncorrelated with their structural agreement and that the verbalized cases reveal qualitatively distinct types of disagreement, including divergences that mirror genuine controversies in the legal commentary.

Shanshan Xu, T. Y. S. S Santosh, Oana Ichim et al.

In legal NLP, Case Outcome Classification (COC) must not only be accurate but also trustworthy and explainable. Existing work in explainable COC has been limited to annotations by a single expert. However, it is well-known that lawyers may disagree in their assessment of case facts. We hence collect a novel dataset RAVE: Rationale Variation in ECHR1, which is obtained from two experts in the domain of international human rights law, for whom we observe weak agreement. We study their disagreements and build a two-level task-independent taxonomy, supplemented with COC-specific subcategories. To our knowledge, this is the first work in the legal NLP that focuses on human label variation. We quantitatively assess different taxonomy categories and find that disagreements mainly stem from underspecification of the legal context, which poses challenges given the typically limited granularity and noise in COC metadata. We further assess the explainablility of SOTA COC models on RAVE and observe limited agreement between models and experts. Overall, our case study reveals hitherto underappreciated complexities in creating benchmark datasets in legal NLP that revolve around identifying aspects of a case's facts supposedly relevant to its outcome.

Shanshan Xu, Leon Staufer, T. Y. S. S Santosh et al. · cambridge

Recognizing vulnerability is crucial for understanding and implementing targeted support to empower individuals in need. This is especially important at the European Court of Human Rights (ECtHR), where the court adapts Convention standards to meet actual individual needs and thus ensures effective human rights protection. However, the concept of vulnerability remains elusive at the ECtHR and no prior NLP research has dealt with it. To enable future research in this area, we present VECHR, a novel expert-annotated multi-label dataset comprising of vulnerability type classification and explanation rationale. We benchmark the performance of state-of-the-art models on VECHR from both prediction and explainability perspectives. Our results demonstrate the challenging nature of the task with lower prediction performance and limited agreement between models and experts. Further, we analyze the robustness of these models in dealing with out-of-domain (OOD) data and observe overall limited performance. Our dataset poses unique challenges offering significant room for improvement regarding performance, explainability, and robustness.

Max Prior, Niklas Wais, Matthias Grabmair

We present a fully automated pipeline that transforms large collections of court decisions into legal commentaries for statutes - without providing any handcrafted doctrinal framework. Using 4.555 decisions of the German Federal Court of Justice that cite sections 242, 280, 812 and 823 of the German Civil Code (BGB), we extract paragraph-level chunks, summarize their reasoning, and derive keywords, which are embedded and clustered. For each cluster, an LLM generates headings and synthesizes citation-rich sections, which are then merged into coherent commentaries by four state-of-the-art LLMs. We evaluate along five dimensions - topical relevance, heading-match, citation faithfulness, cluster distinction and logical ordering - using both a human expert and an LLM-judge. Our results show that commentary-like argument mining from court decisions to generate reports that can be refreshed within minutes at minimal cost is feasible, yet they highlight limitations arising from restricted sources and the normativity of legal reasoning.

T. Y. S. S. Santosh, Mohamed Hesham Elganayni, Stanisław Sójka et al.

Inspired by the legal doctrine of stare decisis, which leverages precedents (prior cases) for informed decision-making, we explore methods to integrate them into LJP models. To facilitate precedent retrieval, we train a retriever with a fine-grained relevance signal based on the overlap ratio of alleged articles between cases. We investigate two strategies to integrate precedents: direct incorporation at inference via label interpolation based on case proximity and during training via a precedent fusion module using a stacked-cross attention model. We employ joint training of the retriever and LJP models to address latent space divergence between them. Our experiments on LJP tasks from the ECHR jurisdiction reveal that integrating precedents during training coupled with joint training of the retriever and LJP model, outperforms models without precedents or with precedents incorporated only at inference, particularly benefiting sparser articles.

Abhishek Agarwal, Shanshan Xu, Matthias Grabmair

Summarizing legal decisions requires the expertise of law practitioners, which is both time- and cost-intensive. This paper presents techniques for extractive summarization of legal decisions in a low-resource setting using limited expert annotated data. We test a set of models that locate relevant content using a sequential model and tackle redundancy by leveraging maximal marginal relevance to compose summaries. We also demonstrate an implicit approach to help train our proposed models generate more informative summaries. Our multi-task learning model variant leverages rhetorical role identification as an auxiliary task to further improve the summarizer. We perform extensive experiments on datasets containing legal decisions from the US Board of Veterans' Appeals and conduct quantitative and expert-ranked evaluations of our models. Our results show that the proposed approaches can achieve ROUGE scores vis-à-vis expert extracted summaries that match those achieved by inter-annotator comparison.

Max Prior, Andreas Schultz, Matthias Grabmair

Large language models are increasingly used for legal research, yet their fixed training cutoffs and reliance on static parametric knowledge are at odds with the evolving nature of statutory law. We study two temporal failure modes: post-cutoff staleness, where models apply superseded rules after legislative amendments, and recency bias, where models prefer newer provisions even when a historical version governs the fact pattern. To this end, we present a benchmark of 312 expert-validated, time-sensitive German statutory QA pairs spanning three categories: Post-Cutoff Amendment Questions, Pre-Amendment Questions, and Multi-Provision Pre-Amendment Questions. We evaluate five LLMs by OpenAI, Anthropic and DeepSeek under four inference settings: Vanilla, Web-search, and two retrieval-augmented variants that enforce temporal validity via a fact date extraction and version filtering. Using an LLM-as-a-judge validated against human expert ratings, we find severe degradation in the Vanilla post-cutoff setting. Both RAG approaches substantially improve performance across all question types, while web search yields unstable gains and exhibits a marked recency bias on historically anchored tasks. Our results indicate that reliable legal QA requires treating temporal validity as a hard constraint.

T. Y. S. S. Santosh, Apolline Isaia, Shiyu Hong et al.

Rhetorical Role Labeling (RRL) of legal documents is pivotal for various downstream tasks such as summarization, semantic case search and argument mining. Existing approaches often overlook the varying difficulty levels inherent in legal document discourse styles and rhetorical roles. In this work, we propose HiCuLR, a hierarchical curriculum learning framework for RRL. It nests two curricula: Rhetorical Role-level Curriculum (RC) on the outer layer and Document-level Curriculum (DC) on the inner layer. DC categorizes documents based on their difficulty, utilizing metrics like deviation from a standard discourse structure and exposes the model to them in an easy-to-difficult fashion. RC progressively strengthens the model to discern coarse-to-fine-grained distinctions between rhetorical roles. Our experiments on four RRL datasets demonstrate the efficacy of HiCuLR, highlighting the complementary nature of DC and RC.

T. Y. S. S. Santosh, Philipp Bock, Matthias Grabmair

Segmentation and Rhetorical Role Labeling of legal judgements play a crucial role in retrieval and adjacent tasks, including case summarization, semantic search, argument mining etc. Previous approaches have formulated this task either as independent classification or sequence labeling of sentences. In this work, we reformulate the task at span level as identifying spans of multiple consecutive sentences that share the same rhetorical role label to be assigned via classification. We employ semi-Markov Conditional Random Fields (CRF) to jointly learn span segmentation and span label assignment. We further explore three data augmentation strategies to mitigate the data scarcity in the specialized domain of law where individual documents tend to be very long and annotation cost is high. Our experiments demonstrate improvement of span-level prediction metrics with a semi-Markov CRF model over a CRF baseline. This benefit is contingent on the presence of multi sentence spans in the document.

T. Y. S. S. Santosh, Irtiza Chowdhury, Shanshan Xu et al.

In high-stakes decision-making tasks within legal NLP, such as Case Outcome Classification (COC), quantifying a model's predictive confidence is crucial. Confidence estimation enables humans to make more informed decisions, particularly when the model's certainty is low, or where the consequences of a mistake are significant. However, most existing COC works prioritize high task performance over model reliability. This paper conducts an empirical investigation into how various design choices including pre-training corpus, confidence estimator and fine-tuning loss affect the reliability of COC models within the framework of selective prediction. Our experiments on the multi-label COC task, focusing on European Court of Human Rights (ECtHR) cases, highlight the importance of a diverse yet domain-specific pre-training corpus for better calibration. Additionally, we demonstrate that larger models tend to exhibit overconfidence, Monte Carlo dropout methods produce reliable confidence estimates, and confident error regularization effectively mitigates overconfidence. To our knowledge, this is the first systematic exploration of selective prediction in legal NLP. Our findings underscore the need for further research on enhancing confidence measurement and improving the trustworthiness of models in the legal domain.

Mohamed Hesham Elganayni, Runsheng Chen, Sebastian Nagl et al.

This work explores the role of prompt design and judge selection in LLM-as-a-Judge evaluations of free text legal question answering. We examine whether automatic task prompt optimization improves over human-centered design, whether optimization effectiveness varies by judge feedback style, and whether optimized prompts transfer across judges. We systematically address these questions on the LEXam benchmark by optimizing task prompts using the ProTeGi method with feedback from two judges (Qwen3-32B, DeepSeek-V3) across four task models, and then testing cross-judge transfer. Automatic optimization consistently outperforms the baseline, with lenient judge feedback yielding higher and more consistent gains than strict judge feedback. Prompts optimized with lenient feedback transfer better to strict judges than the reverse direction. Analysis reveals that lenient judges provide permissive feedback, yielding prompts with broader applicability, whereas strict judges produce restrictive feedback, leading to judge-specific overfitting. Our findings demonstrate algorithmically optimizing prompts on training data can outperform human-centered prompt design and that judges' dispositions during optimization shape prompt generalizability. Code and optimized prompts are available at https://github.com/TUMLegalTech/icail2026-llm-judge-gaming.

Shanshan Xu, Irina Broda, Rashid Haddad et al.

Recent work has demonstrated that natural language processing techniques can support consumer protection by automatically detecting unfair clauses in the Terms of Service (ToS) Agreement. This work demonstrates that transformer-based ToS analysis systems are vulnerable to adversarial attacks. We conduct experiments attacking an unfair-clause detector with universal adversarial triggers. Experiments show that a minor perturbation of the text can considerably reduce the detection performance. Moreover, to measure the detectability of the triggers, we conduct a detailed human evaluation study by collecting both answer accuracy and response time from the participants. The results show that the naturalness of the triggers remains key to tricking readers.

T. Y. S. S. Santosh, Cornelius Weiss, Matthias Grabmair

In the evolving NLP landscape, benchmarks serve as yardsticks for gauging progress. However, existing Legal NLP benchmarks only focus on predictive tasks, overlooking generative tasks. This work curates LexSumm, a benchmark designed for evaluating legal summarization tasks in English. It comprises eight English legal summarization datasets, from diverse jurisdictions, such as the US, UK, EU and India. Additionally, we release LexT5, legal oriented sequence-to-sequence model, addressing the limitation of the existing BERT-style encoder-only models in the legal domain. We assess its capabilities through zero-shot probing on LegalLAMA and fine-tuning on LexSumm. Our analysis reveals abstraction and faithfulness errors even in summaries generated by zero-shot LLMs, indicating opportunities for further improvements. LexSumm benchmark and LexT5 model are available at https://github.com/TUMLegalTech/LexSumm-LexT5.

Sebastian Nagl, Mohamed Elganayni, Melanie Pospisil et al.

Official court press releases from Germany's highest courts present and explain judicial rulings to the public, as well as to expert audiences. Prior NLP efforts emphasize technical headnotes, ignoring citizen-oriented communication needs. We introduce CourtPressGER, a 6.4k dataset of triples: rulings, human-drafted press releases, and synthetic prompts for LLMs to generate comparable releases. This benchmark trains and evaluates LLMs in generating accurate, readable summaries from long judicial texts. We benchmark small and large LLMs using reference-based metrics, factual-consistency checks, LLM-as-judge, and expert ranking. Large LLMs produce high-quality drafts with minimal hierarchical performance loss; smaller models require hierarchical setups for long judgments. Initial benchmarks show varying model performance, with human-drafted releases ranking highest.

T. Y. S. S Santosh, Mahmoud Aly, Matthias Grabmair

Legal professionals frequently encounter long legal judgments that hold critical insights for their work. While recent advances have led to automated summarization solutions for legal documents, they typically provide generic summaries, which may not meet the diverse information needs of users. To address this gap, we introduce LexAbSumm, a novel dataset designed for aspect-based summarization of legal case decisions, sourced from the European Court of Human Rights jurisdiction. We evaluate several abstractive summarization models tailored for longer documents on LexAbSumm, revealing a challenge in conditioning these models to produce aspect-specific summaries. We release LexAbSum to facilitate research in aspect-based summarization for legal domain.

T. Y. S. S Santosh, Tuan-Quang Vuong, Matthias Grabmair

This study investigates the challenges posed by the dynamic nature of legal multi-label text classification tasks, where legal concepts evolve over time. Existing models often overlook the temporal dimension in their training process, leading to suboptimal performance of those models over time, as they treat training data as a single homogeneous block. To address this, we introduce ChronosLex, an incremental training paradigm that trains models on chronological splits, preserving the temporal order of the data. However, this incremental approach raises concerns about overfitting to recent data, prompting an assessment of mitigation strategies using continual learning and temporal invariant methods. Our experimental results over six legal multi-label text classification datasets reveal that continual learning methods prove effective in preventing overfitting thereby enhancing temporal generalizability, while temporal invariant methods struggle to capture these dynamics of temporal shifts.

Santosh T. Y. S. S, Nina Baumgartner, Matthias Stürmer et al.

The assessment of explainability in Legal Judgement Prediction (LJP) systems is of paramount importance in building trustworthy and transparent systems, particularly considering the reliance of these systems on factors that may lack legal relevance or involve sensitive attributes. This study delves into the realm of explainability and fairness in LJP models, utilizing Swiss Judgement Prediction (SJP), the only available multilingual LJP dataset. We curate a comprehensive collection of rationales that `support' and `oppose' judgement from legal experts for 108 cases in German, French, and Italian. By employing an occlusion-based explainability approach, we evaluate the explainability performance of state-of-the-art monolingual and multilingual BERT-based LJP models, as well as models developed with techniques such as data augmentation and cross-lingual transfer, which demonstrated prediction performance improvement. Notably, our findings reveal that improved prediction performance does not necessarily correspond to enhanced explainability performance, underscoring the significance of evaluating models from an explainability perspective. Additionally, we introduce a novel evaluation framework, Lower Court Insertion (LCI), which allows us to quantify the influence of lower court information on model predictions, exposing current models' biases.

T. Y. S. S Santosh, Vatsal Venkatkrishna, Saptarshi Ghosh et al.

Legal professionals face the challenge of managing an overwhelming volume of lengthy judgments, making automated legal case summarization crucial. However, prior approaches mainly focused on training and evaluating these models within the same jurisdiction. In this study, we explore the cross-jurisdictional generalizability of legal case summarization models.Specifically, we explore how to effectively summarize legal cases of a target jurisdiction where reference summaries are not available. In particular, we investigate whether supplementing models with unlabeled target jurisdiction corpus and extractive silver summaries obtained from unsupervised algorithms on target data enhances transfer performance. Our comprehensive study on three datasets from different jurisdictions highlights the role of pre-training in improving transfer performance. We shed light on the pivotal influence of jurisdictional similarity in selecting optimal source datasets for effective transfer. Furthermore, our findings underscore that incorporating unlabeled target data yields improvements in general pre-trained models, with additional gains when silver summaries are introduced. This augmentation is especially valuable when dealing with extractive datasets and scenarios featuring limited alignment between source and target jurisdictions. Our study provides key insights for developing adaptable legal case summarization systems, transcending jurisdictional boundaries.

T. Y. S. S Santosh, Hassan Sarwat, Ahmed Abdou et al.

Rhetorical Role Labeling (RRL) of legal judgments is essential for various tasks, such as case summarization, semantic search and argument mining. However, it presents challenges such as inferring sentence roles from context, interrelated roles, limited annotated data, and label imbalance. This study introduces novel techniques to enhance RRL performance by leveraging knowledge from semantically similar instances (neighbours). We explore inference-based and training-based approaches, achieving remarkable improvements in challenging macro-F1 scores. For inference-based methods, we explore interpolation techniques that bolster label predictions without re-training. While in training-based methods, we integrate prototypical learning with our novel discourse-aware contrastive method that work directly on embedding spaces. Additionally, we assess the cross-domain applicability of our methods, demonstrating their effectiveness in transferring knowledge across diverse legal domains.

Shanshan Xu, T. Y. S. S Santosh, Oana Ichim et al.

In legal decisions, split votes (SV) occur when judges cannot reach a unanimous decision, posing a difficulty for lawyers who must navigate diverse legal arguments and opinions. In high-stakes domains, understanding the alignment of perceived difficulty between humans and AI systems is crucial to build trust. However, existing NLP calibration methods focus on a classifier's awareness of predictive performance, measured against the human majority class, overlooking inherent human label variation (HLV). This paper explores split votes as naturally observable human disagreement and value pluralism. We collect judges' vote distributions from the European Court of Human Rights (ECHR), and present SV-ECHR, a case outcome classification (COC) dataset with SV information. We build a taxonomy of disagreement with SV-specific subcategories. We further assess the alignment of perceived difficulty between models and humans, as well as confidence- and human-calibration of COC models. We observe limited alignment with the judge vote distribution. To our knowledge, this is the first systematic exploration of calibration to human judgements in legal NLP. Our study underscores the necessity for further research on measuring and enhancing model calibration considering HLV in legal decision tasks.

T. Y. S. S. Santosh, Isaac Misael Olguín Nolasco, Matthias Grabmair

Prior case retrieval (PCR) is crucial for legal practitioners to find relevant precedent cases given the facts of a query case. Existing approaches often overlook the underlying semantic intent in determining relevance with respect to the query case. In this work, we propose LeCoPCR, a novel approach that explicitly generate intents in the form of legal concepts from a given query case facts and then augments the query with these concepts to enhance models understanding of semantic intent that dictates relavance. To overcome the unavailability of annotated legal concepts, we employ a weak supervision approach to extract key legal concepts from the reasoning section using Determinantal Point Process (DPP) to balance quality and diversity. Experimental results on the ECtHR-PCR dataset demonstrate the effectiveness of leveraging legal concepts and DPP-based key concept extraction.

T. Y. S. S Santosh, Elvin Quero Hernandez, Matthias Grabmair

Legal professionals often grapple with navigating lengthy legal judgements to pinpoint information that directly address their queries. This paper focus on this task of extracting relevant paragraphs from legal judgements based on the query. We construct a specialized dataset for this task from the European Court of Human Rights (ECtHR) using the case law guides. We assess the performance of current retrieval models in a zero-shot way and also establish fine-tuning benchmarks using various models. The results highlight the significant gap between fine-tuned and zero-shot performance, emphasizing the challenge of handling distribution shift in the legal domain. We notice that the legal pre-training handles distribution shift on the corpus side but still struggles on query side distribution shift, with unseen legal queries. We also explore various Parameter Efficient Fine-Tuning (PEFT) methods to evaluate their practicality within the context of information retrieval, shedding light on the effectiveness of different PEFT methods across diverse configurations with pre-training and model architectures influencing the choice of PEFT method.

T. Y. S. S Santosh, Kristina Kaiser, Matthias Grabmair

In this paper, we introduce CuSINeS, a negative sampling approach to enhance the performance of Statutory Article Retrieval (SAR). CuSINeS offers three key contributions. Firstly, it employs a curriculum-based negative sampling strategy guiding the model to focus on easier negatives initially and progressively tackle more difficult ones. Secondly, it leverages the hierarchical and sequential information derived from the structural organization of statutes to evaluate the difficulty of samples. Lastly, it introduces a dynamic semantic difficulty assessment using the being-trained model itself, surpassing conventional static methods like BM25, adapting the negatives to the model's evolving competence. Experimental results on a real-world expert-annotated SAR dataset validate the effectiveness of CuSINeS across four different baselines, demonstrating its versatility.

T. Y. S. S Santosh, Mahmoud Aly, Oana Ichim et al.

Analyzing large volumes of case law to uncover evolving legal principles, across multiple cases, on a given topic is a demanding task for legal professionals. Structured topical reports provide an effective solution by summarizing key issues, principles, and judgments, enabling comprehensive legal analysis on a particular topic. While prior works have advanced query-based individual case summarization, none have extended to automatically generating multi-case structured reports. To address this, we introduce LexGenie, an automated LLM-based pipeline designed to create structured reports using the entire body of case law on user-specified topics within the European Court of Human Rights jurisdiction. LexGenie retrieves, clusters, and organizes relevant passages by topic to generate a structured outline and cohesive content for each section. Expert evaluation confirms LexGenie's utility in producing structured reports that enhance efficient, scalable legal analysis.

T. Y. S. S. Santosh, Chen Jia, Patrick Goroncy et al.

This paper addresses the task of legal summarization, which involves distilling complex legal documents into concise, coherent summaries. Current approaches often struggle with content theme deviation and inconsistent writing styles due to their reliance solely on source documents. We propose RELexED, a retrieval-augmented framework that utilizes exemplar summaries along with the source document to guide the model. RELexED employs a two-stage exemplar selection strategy, leveraging a determinantal point process to balance the trade-off between similarity of exemplars to the query and diversity among exemplars, with scores computed via influence functions. Experimental results on two legal summarization datasets demonstrate that RELexED significantly outperforms models that do not utilize exemplars and those that rely solely on similarity-based exemplar selection.

Shanshan Xu, T. Y. S. S Santosh, Yanai Elazar et al.

Recent works have shown that Large Language Models (LLMs) have a tendency to memorize patterns and biases present in their training data, raising important questions about how such memorized content influences model behavior. One such concern is the emergence of political bias in LLM outputs. In this paper, we investigate the extent to which LLMs' political leanings reflect memorized patterns from their pretraining corpora. We propose a method to quantitatively evaluate political leanings embedded in the large pretraining corpora. Subsequently we investigate to whom are the LLMs' political leanings more aligned with, their pretrainig corpora or the surveyed human opinions. As a case study, we focus on probing the political leanings of LLMs in 32 US Supreme Court cases, addressing contentious topics such as abortion and voting rights. Our findings reveal that LLMs strongly reflect the political leanings in their training data, and no strong correlation is observed with their alignment to human opinions as expressed in surveys. These results underscore the importance of responsible curation of training data, and the methodology for auditing the memorization in LLMs to ensure human-AI alignment.

T. Y. S. S. Santosh, Hassan Sarwat, Matthias Grabmair

In this paper, we introduce QABISAR, a novel framework for statutory article retrieval, to overcome the semantic mismatch problem when modeling each query-article pair in isolation, making it hard to learn representation that can effectively capture multi-faceted information. QABISAR leverages bipartite interactions between queries and articles to capture diverse aspects inherent in them. Further, we employ knowledge distillation to transfer enriched query representations from the graph network into the query bi-encoder, to capture the rich semantics present in the graph representations, despite absence of graph-based supervision for unseen queries during inference. Our experiments on a real-world expert-annotated dataset demonstrate its effectiveness.

T. Y. S. S Santosh, Rashid Gustav Haddad, Matthias Grabmair

In common law jurisdictions, legal practitioners rely on precedents to construct arguments, in line with the doctrine of \emph{stare decisis}. As the number of cases grow over the years, prior case retrieval (PCR) has garnered significant attention. Besides lacking real-world scale, existing PCR datasets do not simulate a realistic setting, because their queries use complete case documents while only masking references to prior cases. The query is thereby exposed to legal reasoning not yet available when constructing an argument for an undecided case as well as spurious patterns left behind by citation masks, potentially short-circuiting a comprehensive understanding of case facts and legal principles. To address these limitations, we introduce a PCR dataset based on judgements from the European Court of Human Rights (ECtHR), which explicitly separate facts from arguments and exhibit precedential practices, aiding us to develop this PCR dataset to foster systems' comprehensive understanding. We benchmark different lexical and dense retrieval approaches with various negative sampling strategies, adapting them to deal with long text sequences using hierarchical variants. We found that difficulty-based negative sampling strategies were not effective for the PCR task, highlighting the need for investigation into domain-specific difficulty criteria. Furthermore, we observe performance of the dense models degrade with time and calls for further research into temporal adaptation of retrieval models. Additionally, we assess the influence of different views , Halsbury's and Goodhart's, in practice in ECtHR jurisdiction using PCR task.

T. Y. S. S. Santosh, Youssef Farag, Matthias Grabmair

Legal professionals often struggle with lengthy judgments and require efficient summarization for quick comprehension. To address this challenge, we investigate the need for structured planning in legal case summarization, particularly through event-centric representations that reflect the narrative nature of legal case documents. We propose our framework, CoPERLex, which operates in three stages: first, it performs content selection to identify crucial information from the judgment; second, the selected content is utilized to generate intermediate plans through event-centric representations modeled as Subject-Verb-Object tuples; and finally, it generates coherent summaries based on both the content and the structured plan. Our experiments on four legal summarization datasets demonstrate the effectiveness of integrating content selection and planning components, highlighting the advantages of event-centric plans over traditional entity-centric approaches in the context of legal judgements.

Li Zhang, Matthias Grabmair, Morgan Gray et al.

Case-based reasoning is a cornerstone of U.S. legal practice, requiring professionals to argue about a current case by drawing analogies to and distinguishing from past precedents. While Large Language Models (LLMs) have shown remarkable capabilities, their proficiency in this complex, nuanced form of reasoning needs further investigation. We propose a formal framework that decomposes the process of identifying significant distinctions between cases into three-stage reasoning tasks. Our framework models cases using factual predicates called factors, organizes them into a legal knowledge hierarchy, and defines verifiable rules for identifying distinctions, analyzing their argumentative support, and evaluating their significance. Through comprehensive evaluation of modern reasoning LLMs, we reveal a paradox: while models achieve high accuracy on surface-level reasoning (Task 1), performance degrades on hierarchical reasoning (Task 2: 64.82%-92.09%) and collapses on integrated analysis (Task 3: 11.46%-33.99%). Most strikingly, we find that models consistently expend more computational resources on incorrect responses than correct ones, suggesting that "thinking longer" does not always mean "thinking smarter." Our work provides a methodology for fine-grained analysis of LLM reasoning capabilities in complex domains and reveals fundamental limitations that must be addressed for robust and trustworthy legal AI.

Hyunji Lee, Kevin Chenhao Li, Matthias Grabmair et al.

Prompt optimization aims to systematically refine prompts to enhance a language model's performance on specific tasks. Fairness detection in Terms of Service (ToS) clauses is a challenging legal NLP task that demands carefully crafted prompts to ensure reliable results. However, existing prompt optimization methods are often computationally expensive due to inefficient search strategies and costly prompt candidate scoring. In this paper, we propose a framework that combines Monte Carlo Tree Search (MCTS) with a proxy prompt evaluator to more effectively explore the prompt space while reducing evaluation costs. Experiments demonstrate that our approach achieves higher classification accuracy and efficiency than baseline methods under a constrained computation budget.

T. Y. S. S Santosh, Kevin D. Ashley, Katie Atkinson et al.

Modeling legal reasoning and argumentation justifying decisions in cases has always been central to AI & Law, yet contemporary developments in legal NLP have increasingly focused on statistically classifying legal conclusions from text. While conceptually simpler, these approaches often fall short in providing usable justifications connecting to appropriate legal concepts. This paper reviews both traditional symbolic works in AI & Law and recent advances in legal NLP, and distills possibilities of integrating expert-informed knowledge to strike a balance between scalability and explanation in symbolic vs. data-driven approaches. We identify open challenges and discuss the potential of modern NLP models and methods that integrate

Jaromir Savelka, Hannes Westermann, Karim Benyekhlef et al.

In this paper, we examine the use of multi-lingual sentence embeddings to transfer predictive models for functional segmentation of adjudicatory decisions across jurisdictions, legal systems (common and civil law), languages, and domains (i.e. contexts). Mechanisms for utilizing linguistic resources outside of their original context have significant potential benefits in AI & Law because differences between legal systems, languages, or traditions often block wider adoption of research outcomes. We analyze the use of Language-Agnostic Sentence Representations in sequence labeling models using Gated Recurrent Units (GRUs) that are transferable across languages. To investigate transfer between different contexts we developed an annotation scheme for functional segmentation of adjudicatory decisions. We found that models generalize beyond the contexts on which they were trained (e.g., a model trained on administrative decisions from the US can be applied to criminal law decisions from Italy). Further, we found that training the models on multiple contexts increases robustness and improves overall performance when evaluating on previously unseen contexts. Finally, we found that pooling the training data from all the contexts enhances the models' in-context performance.

Zihan Huang, Charles Low, Mengqiu Teng et al.

Lawyers and judges spend a large amount of time researching the proper legal authority to cite while drafting decisions. In this paper, we develop a citation recommendation tool that can help improve efficiency in the process of opinion drafting. We train four types of machine learning models, including a citation-list based method (collaborative filtering) and three context-based methods (text similarity, BiLSTM and RoBERTa classifiers). Our experiments show that leveraging local textual context improves recommendation, and that deep neural models achieve decent performance. We show that non-deep text-based methods benefit from access to structured case metadata, but deep models only benefit from such access when predicting from context of insufficient length. We also find that, even after extensive training, RoBERTa does not outperform a recurrent neural model, despite its benefits of pretraining. Our behavior analysis of the RoBERTa model further shows that predictive performance is stable across time and citation classes.

Mihir Kale, Aditya Siddhant, Sreyashi Nag et al.

Pre-trained word embeddings are the primary method for transfer learning in several Natural Language Processing (NLP) tasks. Recent works have focused on using unsupervised techniques such as language modeling to obtain these embeddings. In contrast, this work focuses on extracting representations from multiple pre-trained supervised models, which enriches word embeddings with task and domain specific knowledge. Experiments performed in cross-task, cross-domain and cross-lingual settings indicate that such supervised embeddings are helpful, especially in the low-resource setting, but the extent of gains is dependent on the nature of the task and domain. We make our code publicly available.

Bhavya Karki, Fan Hu, Nithin Haridas et al.

In this paper, we describe a dataset and baseline result for a question answering that utilizes web tables. It contains commonly asked questions on the web and their corresponding answers found in tables on websites. Our dataset is novel in that every question is paired with a table of a different signature. In particular, the dataset contains two classes of tables: entity-instance tables and the key-value tables. Each QA instance comprises a table of either kind, a natural language question, and a corresponding structured SQL query. We build our model by dividing question answering into several tasks, including table retrieval and question element classification, and conduct experiments to measure the performance of each task. We extract various features specific to each task and compose a full pipeline which constructs the SQL query from its parts. Our work provides qualitative results and error analysis for each task, and identifies in detail the reasoning required to generate SQL expressions from natural language questions. This analysis of reasoning informs future models based on neural machine learning.

Soumya Wadhwa, Varsha Embar, Matthias Grabmair et al.

In this paper, we investigate the tendency of end-to-end neural Machine Reading Comprehension (MRC) models to match shallow patterns rather than perform inference-oriented reasoning on RC benchmarks. We aim to test the ability of these systems to answer questions which focus on referential inference. We propose ParallelQA, a strategy to formulate such questions using parallel passages. We also demonstrate that existing neural models fail to generalize well to this setting.