Towards Realizing Whole-Brain Computational Models Guided by Cognitive Models (WBCM-CogM)
Summary:
In recent years, there has been a focused effort to develop Whole Brain Computational Models (WBCMs), aiming to represent the entire brain's functions and contribute to creating artificial intelligence with human-level capabilities. WBCMs involve not only neuroscientific but also cognitive models, especially in constructing a cognitive architecture for consistency. Cognitive models enhance interpretability in implementing WBCMs into AI agents, providing insight into thought processes. This approach, resembling human cognition, offers potential psychological reassurance to users. The discussion about the relationship between cognitive models and WBCMs is linked to AI alignment debates, crucial as powerful AI systems develop. The workshop aims to discuss methodologies to realize WBCMs, emphasizing the role of cognitive models.
The topic of interests:
Cognitive model / architecture
Whole-brain computational model / Neural network
AI alignment / Interpretability
Artificial general intelligence
General purpose robotics
Predictive coding
Free energy principle
Emotion and memory
Generative AI / Large-Language Model / GPT Utilization
Experiment-based model construction
Date and location:
Date (WCCI): 2024/6/30 - 2024/7/5
This workshop will be held for one full day on the above dates.
Location: PACIFICO Yokohama
1-1-1 Minato Mirai, Nishi-ku, Yokohama, Japan
Please register with WCCI at the following site before you attend this Workshop.
Important Dates:
Paper Submission Deadline: May 1 → May 17 [Extended]
Please prepare a manuscript for poster presentations as well.
CFP:
Application form:
Please use the form below to make your submission.
Manuscripts should be written according to the following guidelines.
Manuscripts should be prepared using the IEEE template
Pages 1-6, including references (Manuscripts are also available in the form of Extended Abstracts.)
Manuscripts should be documented in English
Please be sure to submit it in PDF format.
We welcome you and your colleagues to present papers and participate in discussions.
Submitted manuscripts will be reviewed by the organizers. Manuscripts that deviate significantly from the purpose of the workshop may be rejected.
📌 Following the workshop, we will organize a special issue in the Frontiers journal.
Submission of papers extending your presentations in the workshop is encouraged.
Submissions of papers not presented at this workshop are also welcome.
Contents and Schedule:
TBD
This workshop will consist of a general presentation session and a discussion session. The general presentation sessions include oral and poster presentations (plus short talks). PREFERENCE will be accepted at the time of application, but which one may change depending on the status of the application.
The discussion will focus on the question: How can we develop socially agreeable, interpretable (explainable) brain-inspired cognitive architectures?
We're looking for some invited speakers from diverse backgrounds in neuroscience, AI, and robotics.
Organizers:
Akira Taniguchi
Ritsumeikan University, Japan
Yoshimasa Tawatsuji
The University of Tokyo, Japan
Junya Morita
Shizuoka University, Japan