The Inaugural NABI Symposium — Computational Modeling of the Hippocampus and Contextual Memory
Event Details
- Date:
- Location: Seoul National University Library of Medicine Woobong Hall
- Audience: NABI members and the SNU community
- Bring: Personal laptop recommended for interactive methods sessions.
- Note: Ensure Wi-Fi or internet access is available in advance.
About NABI
NABI (Natural and Artificial Brain Intelligence) is a research community studying the computational principles of the brain and their implications for artificial intelligence. Our work sits at the intersection of computational neuroscience, NeuroAI, and clinical neuroscience, with a particular interest in how mathematical models of neural systems can inform neurology and psychiatry.
The group primarily consists of students and physicians at Seoul National University College of Medicine and SNU Hospital who share an interest in the scientific interface between brain science and AI.
Our activities include:
- Season-based thematic study programs focused on core topics in computational neuroscience
- Collaborative paper review and discussion sessions
- Annual symposia where members present ideas, models, and research directions
The Inaugural NABI Symposium
The first annual NABI Symposium marks the culmination of a year of collaborative study and discussion among NABI members.
The theme of the 2026 symposium, “Memory in Context,” focuses on the hippocampus and contextual memory from a computational perspective. Rather than viewing the hippocampus merely as a storage system for episodic memories, recent theoretical frameworks suggest that it functions as a system for learning the statistical structure of environments and generating predictive representations of future states.
This symposium explores these ideas through modern computational models and neural data, with particular attention to how contextual information shapes memory formation and retrieval.
Program
Section 1: Topic Review
A conceptual overview of contextual memory and hippocampal computation, summarizing recent theoretical and experimental work that motivates current computational models.
Section 2: Methods Deep Dive
Participants will explore computational models of hippocampal representation through hands-on demonstrations and simulations.
- Successor Representation (SR)
An interactive exploration of how predictive state representations may emerge in neural systems. - Clone-Structured Cognitive Graphs (CSCG)
Practical analysis and simulation of hippocampal neural data using the CSCG framework in a Google Colab environment.
Section 3: Research Proposal
Members present new research ideas and open questions at the interface of neuroscience and AI. The goal of this session is to identify unresolved scientific gaps and discuss how computational modeling can help address them.
Contact
For inquiries: admin@nabilab.org
