Worth a look – this AI model spots hidden signs of irregular heartbeats in regular ECGs, catching risks standard tests miss.
MACHINE LEARNING
-
Image Processing with Convolution Kernels Tutorial
By
–
🎥 New Video Alert 🎥https://t.co/8q8vCArYD6
— Satya Mallick (@LearnOpenCV) 29 janvier 2024
Ever wondered how apps and software like Photoshop apply blurring or sharpening effects to images? The secret lies in convolution kernels.
Our latest video tutorial is designed for anyone interested in image processing, whether… pic.twitter.com/kIQA8JfbdYNew Alert https://
youtube.com/watch?v=o0KDTw
a0cR0
… Ever wondered how apps and software like Photoshop apply blurring or sharpening effects to images? The secret lies in convolution kernels. Our latest video tutorial is designed for anyone interested in image processing, whether -
Evaluating Multimodal LLMs: Generalizability, Trustworthiness and Causality
By
–
From GPT-4 to Gemini and Beyond: Assessing the Landscape of MLLMs on Generalizability, Trustworthiness and Causality through Four Modalities Lu et al.: https://
arxiv.org/abs/2401.15071 #ArtificialIntelligence #ChatGPT #DeepLearning -
GIF explaining how large language models work
By
–
If you were ever wondering how LLMs work, this GIF explains it really well!
— God of Prompt (@godofprompt) 29 janvier 2024
pic.twitter.com/3kpulAe3L4If you were ever wondering how LLMs work, this GIF explains it really well!
-
New Technology Wave: Engineering Intelligence and Life at Scale
By
–
Starting to crash around us is a new wave of technology. This wave is unleashing the power to engineer two universal foundations: a wave of nothing less than intelligence, and life…What if the wave is actually a tsunami?
-
Yann LeCun Revisits 2019 Predictions on LLM Limitations
By
–
Nice to see Yann LeCun 2024 making yet another set of points (re the intractability of predicting video from pixel space and the limits of LLMs) that I made in 2019 and 2020. He’s catching up!
-
Model Merging and Pretraining Boost AI Performance Results
By
–
Beyond just merging, we've boosted performance by continuing to pretrain with a massive dataset. Merging alone doesn't cut it for top-notch results.
-
AI/ML Founder Discusses Hallucinations and Model Limitations
By
–
dude, i built an AI/ML company that I sold to Uber, helping design the core IP, and went through the math leading to hallucinatons and troubles with extrapolation etc in 2001. I am not a nontechie.
-
Neural Networks History: From 1940s to Transformers 2017
By
–
(Depends what you count; neural nets go back to the 40s, backprop to roughly the 1960s, etc, though transformers per se only 2017)