RoD-TAL: A Benchmark for Answering Questions in Romanian Driving License Exams

ACL ARR 2025 July Submission817 Authors

Abstract: The intersection of AI and legal systems presents a growing need for tools that support legal education, particularly in under-resourced languages such as Romanian. In this work, we aim to evaluate the capabilities of Large Language Models (LLMs) and Vision-Language Models (VLMs) in understanding and reasoning about Romanian driving law through textual and visual question-answering tasks. To facilitate this, we introduce RoD-TAL, a novel multimodal dataset comprising Romanian driving test questions, text-based and image-based, alongside annotated legal references and human explanations. We implement and assess retrieval-augmented generation (RAG) pipelines, dense retrievers, and reasoning-optimized models across tasks including Information Retrieval (IR), Question Answering (QA), Visual IR, and Visual QA. Our experiments demonstrate that domain-specific fine-tuning significantly enhances retrieval performance. At the same time, chain-of-thought prompting and specialized reasoning models improve QA accuracy, surpassing the minimum grades required to pass driving exams. However, visual reasoning remains challenging, highlighting the potential and the limitations of applying LLMs and VLMs to legal education.

Paper Type: Long

Research Area: Resources and Evaluation

Research Area Keywords: language resources, datasets for low resource languages, multimodal QA

Contribution Types: Approaches to low-resource settings, Data resources

Languages Studied: romanian

Reassignment Request Area Chair: This is not a resubmission

Reassignment Request Reviewers: This is not a resubmission

A1 Limitations Section: This paper has a limitations section.

A2 Potential Risks: Yes

A2 Elaboration: Risks and Ethical Considerations

B Use Or Create Scientific Artifacts: Yes

B1 Cite Creators Of Artifacts: Yes

B1 Elaboration: 3.2, 4.2, B.1, B.4

B2 Discuss The License For Artifacts: No

B2 Elaboration: The code will be released under MIT. The data used in this work is publicly accessible (e.g., the Romanian legislation).

B3 Artifact Use Consistent With Intended Use: Yes

B3 Elaboration: Existing artifacts: 4.1, 4.2, B.4; created artifacts: 3.1, 3.2

B4 Data Contains Personally Identifying Info Or Offensive Content: N/A

B4 Elaboration: The data does not contain personally identifiable data

B5 Documentation Of Artifacts: Yes

B5 Elaboration: 3, A

B6 Statistics For Data: Yes

B6 Elaboration: 3, A

C Computational Experiments: Yes

C1 Model Size And Budget: Yes

C1 Elaboration: B.1

C2 Experimental Setup And Hyperparameters: Yes

C2 Elaboration: 4, B

C3 Descriptive Statistics: Yes

C3 Elaboration: F

C4 Parameters For Packages: Yes

C4 Elaboration: B, supplemental materials

D Human Subjects Including Annotators: No

D1 Instructions Given To Participants: N/A

D2 Recruitment And Payment: N/A

D3 Data Consent: N/A

D4 Ethics Review Board Approval: N/A

D5 Characteristics Of Annotators: N/A

E Ai Assistants In Research Or Writing: No

E1 Information About Use Of Ai Assistants: N/A

Author Submission Checklist: yes

Submission Number: 817