Gabe Ketron
@gketron.bsky.social
Associate Data Scientist @ Cedars Sinai
Opinions are my own.
Opinions are my own.
Good news, everyone!
My first first-author paper is now published in Azheimer’s Research & Therapy! 🎉
My first first-author paper is now published in Azheimer’s Research & Therapy! 🎉
March 12, 2025 at 12:00 AM
Good news, everyone!
My first first-author paper is now published in Azheimer’s Research & Therapy! 🎉
My first first-author paper is now published in Azheimer’s Research & Therapy! 🎉
Reposted by Gabe Ketron
My new piece on The Paradox of Trust and Transparency in AI www.extremetech.com/science/the-... #artificialintelligence #trust #transparency
The Paradox of Trust and Transparency in AI
An AI is useless if you can't trust it—so how do we get there, and in a way that benefits and protects all parties involved?
www.extremetech.com
March 6, 2025 at 2:57 PM
My new piece on The Paradox of Trust and Transparency in AI www.extremetech.com/science/the-... #artificialintelligence #trust #transparency
Reposted by Gabe Ketron
PLEASE SHARE 🧪‼️
Since CDC websites and datasets are being taken down by the Administration…
The ENTIRE archive of ALL CDC datasets uploaded BEFORE January 28th, 2025 can be found here: archive.org/details/2025...
Since CDC websites and datasets are being taken down by the Administration…
The ENTIRE archive of ALL CDC datasets uploaded BEFORE January 28th, 2025 can be found here: archive.org/details/2025...
February 2, 2025 at 5:00 PM
PLEASE SHARE 🧪‼️
Since CDC websites and datasets are being taken down by the Administration…
The ENTIRE archive of ALL CDC datasets uploaded BEFORE January 28th, 2025 can be found here: archive.org/details/2025...
Since CDC websites and datasets are being taken down by the Administration…
The ENTIRE archive of ALL CDC datasets uploaded BEFORE January 28th, 2025 can be found here: archive.org/details/2025...
Free LLM coding guide!
Do you want to get in on LLM-assisted coding but don’t know where to start? You can set it up for free in minutes by running a small but powerful model on your machine with Ollama, and connecting it to your IDE of choice (as long as it is VScode).
🧵
1/5
Do you want to get in on LLM-assisted coding but don’t know where to start? You can set it up for free in minutes by running a small but powerful model on your machine with Ollama, and connecting it to your IDE of choice (as long as it is VScode).
🧵
1/5
January 24, 2025 at 7:36 PM
Free LLM coding guide!
Do you want to get in on LLM-assisted coding but don’t know where to start? You can set it up for free in minutes by running a small but powerful model on your machine with Ollama, and connecting it to your IDE of choice (as long as it is VScode).
🧵
1/5
Do you want to get in on LLM-assisted coding but don’t know where to start? You can set it up for free in minutes by running a small but powerful model on your machine with Ollama, and connecting it to your IDE of choice (as long as it is VScode).
🧵
1/5
Large Language Models like GPT-4 are reshaping how we tackle knowledge-intensive tasks, enhancing productivity & quality. Unlike tangible outputs in manufacturing or sales, research production is measured by impact—citations, collaborations, and real-world applications.
January 21, 2025 at 11:04 PM
Large Language Models like GPT-4 are reshaping how we tackle knowledge-intensive tasks, enhancing productivity & quality. Unlike tangible outputs in manufacturing or sales, research production is measured by impact—citations, collaborations, and real-world applications.
I developed a Monte Carlo permutation test tailored for small sample sizes where normal distribution isn’t assumed. 📊By leveraging Fisher–Pitman principles, we provide a robust alternative to Wilcoxon tests, ensuring accuracy even with limited, non-normally distributed paired data. #DataScience
January 21, 2025 at 10:42 PM
I developed a Monte Carlo permutation test tailored for small sample sizes where normal distribution isn’t assumed. 📊By leveraging Fisher–Pitman principles, we provide a robust alternative to Wilcoxon tests, ensuring accuracy even with limited, non-normally distributed paired data. #DataScience
Skyrocketing costs begins now. Insulin prices will be 2x or more by the end of the week. The marks the start of a massive trump wealth transfer from the middle and lower classes to the elite. The culture war policies are a distraction from this main event for his donors.
January 21, 2025 at 7:12 PM
Skyrocketing costs begins now. Insulin prices will be 2x or more by the end of the week. The marks the start of a massive trump wealth transfer from the middle and lower classes to the elite. The culture war policies are a distraction from this main event for his donors.
This is my Costco grocery list from yesterday. Day 1, let's see how he does over the next 4 years.
January 21, 2025 at 1:10 AM
This is my Costco grocery list from yesterday. Day 1, let's see how he does over the next 4 years.
Bioengineers and AI researchers, applications are extended to Jan 31st!
We are accepting application through Jan 31 for our new PhD program in Health AI at Cedars-Sinai Medical Center in Los Angeles, CA www.cedars-sinai.edu/education/gr... #artificialintelligence #bioinformatics #datascience #phd #phdchat #gradschool
PhD Program in Health Artificial Intelligence
The PhD program in Health Artificial Intelligence at Cedars-Sinai prepares students with rigorous training in AI algorithms and methods to improve patient care.
www.cedars-sinai.edu
January 21, 2025 at 12:44 AM
Bioengineers and AI researchers, applications are extended to Jan 31st!
Reposted by Gabe Ketron
Biomedical Engineers: Interested in a computer and data science research career? Come join us in Cedars Sinai’s new Health AI Phd Program! www.cedars-sinai.org/discoveries/...
Cedars-Sinai Pioneers a New Era for AI Education
Cedars-Sinai is revolutionizing health education with its Health Sciences University’s new PhD in Health AI Program.
www.cedars-sinai.org
January 10, 2025 at 11:26 PM
Biomedical Engineers: Interested in a computer and data science research career? Come join us in Cedars Sinai’s new Health AI Phd Program! www.cedars-sinai.org/discoveries/...
I'm surprised no one on X or here is talking about a shocking new development coming out of Neurips 2024 last December resulting in 1000x increases to neural network training speed: the Stochastic Taylor Derivative Estimator.
January 13, 2025 at 8:36 PM
I'm surprised no one on X or here is talking about a shocking new development coming out of Neurips 2024 last December resulting in 1000x increases to neural network training speed: the Stochastic Taylor Derivative Estimator.
my thesis advisor found some wild images of the fires happening here in LA
#Sentinel-2 caught the start of the #palisadesfire. Photo was taken 10:45 am this morning. Second photo is a closeup using false color (IR bands). It looks like the fire started right along Temescal Ridge Trail
January 10, 2025 at 11:45 PM
my thesis advisor found some wild images of the fires happening here in LA
god in heaven i need a bluesky tweet scheduler stat
January 10, 2025 at 11:41 PM
god in heaven i need a bluesky tweet scheduler stat
Just finished watching Dwarkesh Patel’s interview with Adam Brown, and his experience with AI tools very much lines up with mine. Using a chatbot (GPT-4o) to help learn niche medium to hard level skills that introductory material may not cover saves me a ton of time and stress.
January 10, 2025 at 11:40 PM
Just finished watching Dwarkesh Patel’s interview with Adam Brown, and his experience with AI tools very much lines up with mine. Using a chatbot (GPT-4o) to help learn niche medium to hard level skills that introductory material may not cover saves me a ton of time and stress.
Did you know that most LLMs forgo encoders and use decoder-only structures? They reference their own output in training, letting you get higher next word prediction accuracy with the same parameters & are Turing complete under the right conditions. arxiv.org/abs/2305.17026
How Powerful are Decoder-Only Transformer Neural Models?
In this article we prove that the general transformer neural model undergirding modern large language models (LLMs) is Turing complete under reasonable assumptions. This is the first work to directly ...
arxiv.org
January 10, 2025 at 11:35 PM
Did you know that most LLMs forgo encoders and use decoder-only structures? They reference their own output in training, letting you get higher next word prediction accuracy with the same parameters & are Turing complete under the right conditions. arxiv.org/abs/2305.17026
Program synthesis and code pairing tools still need human intervention but formal-method synthesis is making a comeback as a potential solution: arxiv.org/abs/2410.19736
Combining LLM Code Generation with Formal Specifications and Reactive Program Synthesis
In the past few years, Large Language Models (LLMs) have exploded in usefulness and popularity for code generation tasks. However, LLMs still struggle with accuracy and are unsuitable for high-risk ap...
arxiv.org
January 10, 2025 at 11:30 PM
Program synthesis and code pairing tools still need human intervention but formal-method synthesis is making a comeback as a potential solution: arxiv.org/abs/2410.19736
Biomedical Engineers: Interested in a computer and data science research career? Come join us in Cedars Sinai’s new Health AI Phd Program! www.cedars-sinai.org/discoveries/...
Cedars-Sinai Pioneers a New Era for AI Education
Cedars-Sinai is revolutionizing health education with its Health Sciences University’s new PhD in Health AI Program.
www.cedars-sinai.org
January 10, 2025 at 11:26 PM
Biomedical Engineers: Interested in a computer and data science research career? Come join us in Cedars Sinai’s new Health AI Phd Program! www.cedars-sinai.org/discoveries/...
Just wrapped up a study to find the best way to pass CSV data as an LLM API query. Did you know that using the equality operator is effective for communicating multi-shot examples to your model? Until multimodal LLMs can accept CSV files, I’d recommend you try something this.
January 10, 2025 at 11:23 PM
Just wrapped up a study to find the best way to pass CSV data as an LLM API query. Did you know that using the equality operator is effective for communicating multi-shot examples to your model? Until multimodal LLMs can accept CSV files, I’d recommend you try something this.
Reposted by Gabe Ketron
Cedars-Sinai Pioneers a New Era for AI Education www.cedars-sinai.org/discoveries/... #artificialintelligence #machinelearning #phd #bioinformatics #gradschool
Cedars-Sinai Pioneers a New Era for AI Education
Cedars-Sinai is revolutionizing health education with its Health Sciences University’s new PhD in Health AI Program.
www.cedars-sinai.org
January 10, 2025 at 5:09 PM
Cedars-Sinai Pioneers a New Era for AI Education www.cedars-sinai.org/discoveries/... #artificialintelligence #machinelearning #phd #bioinformatics #gradschool
anyone have good recommendations for good open source text corpuses? preferably ones without copywrite infringement
November 27, 2024 at 5:39 PM
anyone have good recommendations for good open source text corpuses? preferably ones without copywrite infringement
Cookim in the kitchen
November 27, 2024 at 3:47 AM
Cookim in the kitchen
its not delusional to believe you are being gangstalked, just ask any of my 50 invisible stalkers
November 26, 2024 at 6:52 PM
its not delusional to believe you are being gangstalked, just ask any of my 50 invisible stalkers
Is it cringe to add an MS degree to your bio?
November 26, 2024 at 6:49 PM
Is it cringe to add an MS degree to your bio?