@leonglambert

ML/AL Applied Scientist | Health and Medicine

 Honolulu, HI
lambertleong.com
Joined June 2018
  • Tweets55
  • Following57
  • Followers45
Excited to dive into #agenticAI with @TDataScience ! My latest article explores how to evaluate these solutions vs. traditional models, specific to health and medicine. A key for 2026. #AI #DataScience. Are these agents really better?
Towards Data Science
@TDataScience
30 Dec 2025
Ensure your new agentic solution is a true improvement. @leonglambert's new article teaches you how to generate continuous scores from agent outputs. towardsdatascience.com/agent…

Agents Under the Curve (AUC)

Towards understanding if your agentic solution is actually better

towardsdatascience.com
2
53
Excel is an interface, not data
12
Would love to collaborate with the @xai and @grok team on some medical and health AI application development efforts we have in the works!
1
22
Lambert Leong, PhD retweeted
Towards Data Science
@TDataScience
Jan 27
Learn to calculate a valid AUC for your agentic AI solution. @leonglambert shares 6 methods, from extracting log probabilities to using Monte Carlo sampling, for benchmarking your models correctly. towardsdatascience.com/agent…

Agents Under the Curve (AUC)

Towards understanding if your agentic solution is actually better

towardsdatascience.com
1
12
934
Lambert Leong, PhD retweeted
Towards Data Science
@TDataScience
Jan 17
Agentic systems often give yes-or-no answers, but how do you evaluate them properly? @leonglambert explains how to make AUC work for agentic AI without breaking evaluation standards. towardsdatascience.com/agent…

Agents Under the Curve (AUC)

Towards understanding if your agentic solution is actually better

towardsdatascience.com
2
6
1,057
Agentic AI in Healthcare is gaining steam but we need to be sure there is actual value add
Towards Data Science
@TDataScience
Jan 15
AUC is still the gold standard in healthcare modeling, but agentic outputs don’t fit neatly. @leonglambert walks through practical ways to turn agent decisions into meaningful continuous scores. towardsdatascience.com/agent…

Agents Under the Curve (AUC)

Towards understanding if your agentic solution is actually better

towardsdatascience.com
1
29
Lambert Leong, PhD retweeted
Towards Data Science
@TDataScience
Jan 13
Struggling to benchmark your agentic system against a standard ML model? @leonglambert's article includes 6 ways to create continuous scores from agentic outputs for a fair AUC-based evaluation. towardsdatascience.com/agent…

Agents Under the Curve (AUC)

Towards understanding if your agentic solution is actually better

towardsdatascience.com
1
3
981
Lambert Leong, PhD retweeted
Towards Data Science
@TDataScience
Jan 4
.@leonglambert dives into actionable techniques for generating continuous scores, a prerequisite for the AUC evaluations that clinicians and reviewers expect. towardsdatascience.com/agent…

Agents Under the Curve (AUC)

Towards understanding if your agentic solution is actually better

towardsdatascience.com
1
1
3
1,548
Some thoughts I had while trying to understand if all these agentic solutions are better pub.towardsai.net/agents-und…

Agents Under the Curve (AUC)

Towards understanding if your agentic solution is actually better

pub.towardsai.net
21
Just had some PhD graduates who admitted to me that they had data sitting on a cluster waiting to be analyzed. 🤯 They never have to source even from a public repo or clean their data... it shows
22
During my computer science graduate program, no one mentioned, taught, or referenced the scientific method
21
Year books are a valuable source of AI training data. The images and the labels are right there
18
Do you remember when you joined X? I do! #MyXAnniversary
15
If research is conducted with tax money, NIH or NSF or etc., why do we have to pay journals to read the paper talking about the research?
1
15
The CGI in AI Gen movies will be to fix the text, fingers, and hair
30
Starting to believe that the path to AGI is not through LLMs
1
37
I know transformers are the current state of the art but people underestimate the use of auto encoders in industry
29
NEW: #FDA cleared #AI devices! linkedin.com/posts/lambert-l…
1
33
If we can be an organ donor, can we be a data donor? #Healthcare #ai #healthai
20
Attempting to push the boundaries of generative AI for medical imaging with quantitative accuracy 🚀 Thrilled to share our work highlighted by @TDataScience . Dive into how we're using 3D scans and #GenAI to predict body composition like never before. #AI #medicalimaging
Towards Data Science
@TDataScience
8 Feb 2024
Learn how a new generative medical-imaging model can predict a body's interior composition based on 3D images — @leonglambert walks us through the latest research he and his coauthors recently published. buff.ly/4bp85Y7
4
4
3,243
Load more