Doctoral Researcher in Multimodal and Multilingual Language Models

We offer a fully funded PhD student position within the TruX Research Group headed by Prof. Dr. Tegawendé F. Bissyandé. The position is supported by Google.org and is embedded in a broader research agenda on multimodal and multilingual language models. The successful candidate will focus on the development of multimodal, multilingual resources and on the design and evaluation of methods that leverage multimodal signals to improve the performance of multilingual models for low-resource languages, such as Luxembourgish. Particular emphasis will be placed on language-agnostic modalities, including images and speech, as bridges to enhance cross-lingual transfer between high- and low-resource languages. The project combines resource creation with methodological research and aims to advance inclusive, robust, and scalable multilingual AI systems.

TruX conducts research in software security, software repair, and explainable software to create key practical solutions for developers, allowing them to achieve trustworthiness, efficiency, and transparency. Application areas include FinTech, embedded systems (e.g., mobile), business or entertainment systems, cybersecurity, and more. For more information, you may refer to https://www.uni.lu/snt-en/research-groups/trux/ .

The successful candidate will:

• Conduct cutting-edge research in multimodal and multilingual natural language processing
• Develop and curate multimodal, multilingual resources for low-resource languages
• Design, implement, and evaluate methods that leverage multimodal signals, such as images and speech, to improve multilingual and cross-lingual language models
• Contribute to national and international research collaborations
• Prepare project deliverables and disseminate results through high-quality scientific publications
• Present research findings at leading international conferences and workshops
• Participate in the supervision and mentoring of MSc and BSc students working on related topics
• Support the organization of relevant demonstrations, shared tasks, and workshops

For further information, please get in touch with Dr. Jordan Samhi (email address: jordan.samhi@uni.lu ). Use the job title and reference number in the subject line of your email.