Not all of my fine-tunes work perfectly
MACHINE LEARNING
-
Advocating for Machine-Generated Data in AI Learning
By
–
Humans largely learn from human-generated data. Machines currently learn from human-generated data. It makes sense for machines to learn from machine-generated data. x.com/peterjliu/stat…
-
MedAlign Dataset Wins Best Generative AI Health Paper Award
By
–
Thrilled to announce: "MedAlign: A Clinician-Generated Dataset for Instruction Following with Electronic Medical Records" just won the "Best Findings Paper in Generative AI for Health" at the ML4H Symposium! Authored by Scott Fleming, Alejandro Lozano, Snorkel AI
-
AI Advances Climate Modeling for Stratospheric Aerosol Injection Research
By
–
Had lunch with a group of distinguished climate scientists to discuss applying AI to climate modeling to advance the science of stratospheric aerosol injection (SAI) for cooling Earth. AI can help figure out if this is actually a good idea (which we do not currently know).
-
Di Wang Wins SODA 2024 Best Paper Award Minimum Cut
By
–
Congrats to Di Wang & coauthors for co-winning the SODA 2024 Best Paper Award for "Deterministic Near-Linear Time Minimum Cut in Weighted Graphs", which is a significant development since Karger's famous randomized algorithm for this problem. Learn more→ https://
goo.gle/48mXpqr -
Neural Network Resilience Through Summation and Dense Connectivity
By
–
I follow the summations and network connectivity to understand resilience to information loss.
When each output is the sum over a large set of input computations, which themselves densely cross-connect in previous layers — I just expect information to be really resilient to -
UK Partners Globally to Accelerate Development Using AI
By
–
UK unites with global partners to accelerate development using AI https://
bit.ly/3t56F3C
#AI #MachineLearning #DeepLearning #LLMs #DataScience -
Language Learning Easier Than Intuitive Physics Prediction
By
–
It was quite obvious 5-10 years ago that learning the structure of language by predicting missing words was easier than learning, say, intuitive physics by predicting video.
We have had projects on both things at FAIR for almost 10 years, and the language side made progress much -
Why LLMs Pass Bar Exams But Robots Cannot Clear Dinner Tables
By
–
Roboticists and computer vision people have all made this point, including Moravec, Rodney Brooks, Jitendra Malik: intelligence requires sensory input. Why do we have LLMs that can pass the bar exam, but no robot that can clear up the dinner table and fill up the dishwasher, a
-
Simplifying Complex AI/ML Concepts: What Should I Explain?
By
–
I think I'm going to make more videos where I try to simplify more complex topics in the world of AI/ML. What terms or concepts have you heard thrown around that you could use a better explanation of? Are there any models or systems that you have no clue how they work? I really