2025.09.16
Events & Symposiums
AGIS

Japan Scientist AI Jam Session 2025
(Application Deadline: Friday, October 24)

RIKEN is pleased to announce an event where participants will engage with cutting-edge AI models from Amazon Web Services, Anthropic, Google, and OpenAI, and apply them to tackle specific “challenging problems” in scientific research.

What is a “Jam Session”?

The term originated in the music field, referring to musicians playing improvisationally without predetermined songs or arrangements.
When used in the AI field, it embodies the spirit of this jam session: “Scientists gather to brainstorm ideas spontaneously using the latest AI models, collaboratively producing new research outcomes.”
In the United States, the Department of Energy, OpenAI, and Anthropic jointly hosted the “1000 Scientist AI Jam Session” on February 28, 2025, with over 1,500 scientists participating. RIKEN is organizing the Jam Session for Japanese scientists.

Participating Companies

  • Amazon Web Services
  • Anthropic
  • Google
  • OpenAI
  • (Alphabetical order)

Schedule

Online Tutorial
  • Date and Time: Tuesday, November 11, 2025, 9:00 AM to 12:00 PM
  • Venue: Online (Zoom Webinar)
  • Content: This session will explain the purpose of this event, the basic operation of the latest large language model (LLM) AI models, and examples of their application to research topics. It serves as a preparatory step for participants to engage with the latest AI models and develop the skills to utilize them.
Dry Run
  • Date and Time: Friday, November 28, 2025, 9:00 AM to 5:00 PM
  • Venue: RIKEN Tokyo Liaison Office and RIKEN Kobe R-CCS
  • Content: We will conduct operational checks of systems and environments, as well as simulated trials, in preparation for the main session. Participants can advance their preparations for utilizing AI models in an environment that closely mirrors the actual session.
Day 1: Main Session
  • Date and Time: December 16, 2025, 9:00 AM to 5:00 PM
  • Venue: AP Nihonbashi (*in Japanese only)(Nihonbashi Front, 3-6-2 Nihonbashi, Chuo-ku, Tokyo)
  • Content: Participants will access cutting-edge AI models from Amazon Web Services, Anthropic, Google, and OpenAI, and conduct hands-on experiments to explore how AI can be applied to tackle specific “challenging problems” in their scientific research fields. Staff from AI companies, including Anthropic, will be on-site to provide technical support.
Day 2: Main Session
  • Date and Time: Thursday, December 18, 2025, 9:00 AM - 5:00 PM
  • Venue: RIKEN Kobe R-CCS
  • Content: Continuing from Day 1, we will explore the potential of AI utilization in scientific research. This is a crucial opportunity to evaluate model responses and provide feedback to AI companies to improve future AI systems according to scientists' needs. Staff from AI companies, including Anthropic, will be on-site to provide support.

*Tutorials are online only. All other sessions are on-site only.
*You can participate on multiple days, including Day 1 and Day 2.

Participation Requirements

  • Must be a scientist (including graduate students and those involved in research management or research communication)
  • Your affiliated institution must sign an NDA with Google and others (RIKEN, as the organizer, will sign on behalf of participants)
  • You must submit the prompts and data used later
  • You must be able to bring your own laptop

*To prioritize researchers whose work aligns with the event's objectives, we will review applications after registration. If applications exceed capacity, participants will be selected by lottery regardless of research topic. Please note that the details of the review and lottery process cannot be disclosed.

Data and Log Provision

The data and logs you provide will be screened by RIKEN and other participating organizations before being provided to the AI company. We appreciate your cooperation.

Application Form

Please apply using the form below. [Application Deadline: Friday, October 24, 2025]
https://forms.gle/jKpDNYPTu1ydhJkV6

Contact Information

Click to display email address

*Please note that the content on this page is updated periodically.

pagetop