Workshop "Large Language Models (LLMs) in Brain Research"
The 1.5-day workshop on "Large Language Models (LLMs) in Brain Research" on October 27 and 28, 2025, brought together leading researchers in computational cognitive neuroscience and artificial intelligence. Elaheh Akbari and Giacomo Aldegheri have organized the workshop with support of the Giessen Graduate Center for Natural Sciences and Psychology (GGN). The event fostered rich interdisciplinary dialogue, facilitated meaningful collaborations, and advanced collective understanding of how LLMs and vision-language models relate to brain function, neural representation, and cognitive processes.
|
The workshop featured cutting-edge talks from renowned experts including Prof. Dr. Adrien Doerig, Dr. Sandro Pezzelle, and Prof. Mariya Toneva, Ph.D., who addressed diverse topics including model-brain alignment, neural circuit mechanisms, multimodal representations, and practical applications of LLMs in experimental neuroscience. Speakers presented novel frameworks for comparing computational representations to brain activity, discussed decoding visual scenes from neural signals, and outlined brain-inspired approaches to improving model alignment with neural data. The breadth of topics—from circuits and representations to behavior and language processing—created a rich intellectual environment. |
|
The workshop successfully achieved its core objectives: Participants gained conceptual grounding and practical skills for leveraging LLMs in experimental design and neural data analysis; established connections across disciplines that are poised to generate future collaborations; and identified key open questions advancing the field of brain-inspired AI. The combination of rigorous scientific content with accessible instruction created an environment where both early-career and established researchers could engage meaningfully. The sustained engagement and quality of discourse reflected the vibrant, growing community committed to bridging neuroscience and artificial intelligence. Conversations between speakers and attendees ranged across cutting-edge topics including mechanistic alignment between LLMs and neural circuits, and the interpretability of model representations, methodological best practices for brain-model comparisons. Speakers showed a genuine collaborative spirit, building connections across presentations. |
 |
The workshop demonstrated that the cognitive neuroscience and AI communities are primed for deeper integration, with LLMs serving as a powerful focal point for collaborative inquiry. The fruitful conversations, expert talks, and engaged audience have positioned this gathering as a meaningful contribution to advancing our understanding of how computational models illuminate brain function.