CAIS 2025: Call for Tutorials

Call for Tutorials

CAIS 2025 calls for tutorials that could provide background knowledge and/or great insights into both past experiences, shared challenges, and general methodologies that many AI for science fields would benefit from.

Goal and Audience

The goal of the CAIS 2025 tutorials is to introduce general methods and share interesting problems in AI for sciences from an educational perspective. The audience of the tutorials should be experts and students from both machine learning and related science fields.

Submission link: Please submit your proposal in PDF or Word format using our Google form: https://docs.google.com/forms/d/1Ygt6mclH3dnfT9NnieXaqpKpe2Qu4QKhNb0CB4wf5G8/edit

Tutorial proposal submission timeline (Also see Dates) All deadlines are 23:59 AoE (anywhere on earth)

  • Tutorial proposal deadline: May 9, 2025

  • Proposal decision notifications: June 2, 2025

CAIS 2025 tutorial committee

The CAIS 2025 tutorial committee will commit to providing support for each tutorial and coordinate all tutorial events. Here are our members you can reach out via email. (Please also label “CAIS 2025 Tutorial” in the title)

Chair: Lixue Cheng (Microsoft Research AI for Science Lab) sherrylixuecheng@gmail.com

Tutorial topics, format and suggestions

The tutorials will be general conference sessions and only have a single event happening in one time slot. The length of each tutorial should be 2.5 - 3 hours. We have 3 topics that have been proposed by the conference organizing teams as example tutorials but need contributions and would like to seek another 3-4 tutorials proposed by submission.

The following are the general and suggested format for the tutorial, but any new idea is welcome and could be discussed with the CAIS 2025 tutorial committee.

Proposal format:

  1. 3-4 page well-written proposal in Neurips paper style.

  2. Clearly state the goals and necessity of having this tutorial.

  3. Have proposed contents which are suitable for 3-4 hours and focus on the approaches and challenges.

CAIS 2025 organizing team proposed example topics:

  1. Tutorial on recent development of generative models and diverse application in Sciences: diffusion/flow matching/stochastic interpolant/action matching/rectified flow from methodological formulation to diverse science applications.

  2. Tutorial on LLMs for Sciences: Methodologies, progresses and challenges: LLMs have rapidly transformed scientific discovery by enabling researchers to efficiently process vast amounts of literature and generate novel insights from complex data sets. A dedicated tutorial on LLMs for Sciences will equip participants with practical skills and a deeper understanding of current available tools, and recent processes of LLMs for Sciences and the key challenges to make it more useful for discovery.

  3. Tutorial on AI for drug discovery: This tutorial might introduce machine learning techniques, real-world case studies, practical applications, and current challenges in AI for drug design.

  4. Other proposed topics: Waiting for your contributions!

Suggested:

Since our audience has diverse backgrounds, for a successful tutorial, we suggest the organizers Consider contents that are educative that could provide a great overview of the approach and problem. Consider topics that have a great potential to be/have been applied to several scientific problems Provide additional reading material lists for further exploration

Tutorial organizer/speaker

Even though the proposal could be submitted by one organizer, we expect each tutorial should have 3-4 organizers/presenters and at least one can present on-site during CAIS 2025 (13-16 October, 2025, Princeton, NJ).

For the accepted tutorial, the organizers need to provide a final outline of the contents to the CAIS 2025 symposia committee before 30 Sep, 2025.