2023-08-02 AUG

Preparing/discussing agenda for Aug 9

We have three workshop/presentation/interactions next Wednesday (Aug 9)

The first, at C4B 9:30-11:30am, led by Wolfram Alpha;
the second at the Bioinformatics Journal Club by Jeya, 12:30-1:30pm;
and at 2-3pm the social hour hosted by TDRP where we can be engaged more informally

We'll use today's C4B Journal Club splot to prepare for it.

9:30am Wolfram Alpha operation with LLMs

We look forward to spending a couple of hours demonstrating Wolfram Technologies within the AI and Large Language Model use space. All in attendance will need a free Cloud Account to do anything hands-on following along.

https://www.wolframcloud.com

https://products.wolframalpha.com/api

Suggested schedule:

First 5 mins - Introductions

Next 30-45 mins - Wolfram Language

Creating and Consuming API's - Both for data and computation

Project results into the embeddings space and use W|A bio data and knowledge

Chat Notebooks

Next 30-45 mins - LLM's

Learning models and how to train or re-train

OpenAPI plugin – show code and what is going on

Show example from related ongoing consulting project with Brooke Army Medical Center - making public cancer datasets fully computable in WL

The remaining minutes we have we'll do Q&A.

Thank you again for the opportunity!

Brad Harden & Stephanie Harpst

12:30pm GPT ( Jeya )

Jeya is presenting GPT next week at the Bioinformatics Journal Club, with demonstrative contributions from others as he sees fit.

Here's the announcement

The next Journal Club meeting (https://nci-dceg.github.io/bioinformatics_journal_club) is scheduled for 12:30-1:30pm, August 9, 2023. The hybrid meeting will take place at Rm 6E032/34/ via Teams (meeting details below). Please take note of this one-time alteration in our usual schedule.

Jeya B. Balasubramanian, postdoctoral fellow in the Data Science Research GroupIn this session, Dr. Jeya Balasubramanian (TDRP) will be leading the discussion on Large Language Models (LLMs) like GPT-4 and AI agents, and how they could be applied in cancer research.

We will do a quick historical review of the field of AI centering around the events that led to the creation of large language models, such as ChatGPT. We will take a close look at what ChatGPT does and assess opportunities and limitations that can arise from it. We will also speculate on the future of AI.

We will be loosely following the following blog entry:

What Is ChatGPT Doing … and Why Does It Work?

Stephen Wolfram

Link: https://writings.stephenwolfram.com/2023/02/what-is-chatgpt-doing-and-why-does-it-work

2:00pm DCEG Social Gathering hosted by TDRP

DCEG Social Gathering hosted by TDRP

Eat donuts and talk to people ;-)

Hackathon

Hands-on review of closing Summer projects and initiatives

...

Notes

...

ISMB2023 highlights (if we have time - check the link for specific topics/questions)

for example, at https://www.iscb.org/ismbeccb2023-programme/nih-odss-elixir see session reporting on federated cloud infrastructure.
why data science, by Phill Bourne https://iscb.junolive.co/ismb2023/live/panels/Special_Session_Block_3_Day_4 > https://globalbiodata.org
...

Bringing TensorFlow back

if we are to keep up with neural computing at all. See for example https://www.nature.com/articles/s42256-023-00668-8

Report abuse