Browsing Tag
deeplearning
25 posts
Deep Learning Without Backpropagation
Most modern neural networks learn using backpropagation. It works well, but it has a strange property: learning depends…
Backprop Finally Made Sense When I Reimplemented It in Rust
I never used PyTorch or TensorFlow. My ML background was NumPy and scikit-learn. I could train models, tune…
Fast Transformer Decoding: One Write-Head is All You Need
Faster Transformer Decoding — One Write-Head Changes How AI Replies Imagine your phone trying to build a sentence…
Deep Graph Library: A Graph-Centric, Highly-Performant Package for Graph NeuralNetworks
Deep Graph Library — Faster, lighter tools for learning from networks Imagine a toolbox that helps computers learn…
SmoothGrad: removing noise by adding noise
SmoothGrad: why adding tiny noise can make picture explanations clearer Have you ever wondered why a computer calls…
Linear Algebra for AI
Linear Algebra for AI — Part 1: What Is Linear Algebra? (The Big Picture) Last updated: 23 Nov…
PICABench: How Far Are We from Physically Realistic Image Editing?
Can AI Make Photo Edits Look Totally Real? Ever wondered why a removed object sometimes leaves a ghostly…
InfiMed-ORBIT: Aligning LLMs on Open-Ended Complex Tasks via Rubric-BasedIncremental Training
How AI Doctors Get Smarter with Rubric Training Ever wondered how a chatbot could give you reliable medical…
MTSQL-R1: Towards Long-Horizon Multi-Turn Text-to-SQL via Agentic Training
AI Agent Learns to Talk to Databases Over Long Conversations Ever wondered how a chatbot could actually fetch…
Very insightful!
DeepSeek-OCR: When a Picture Is Actually Worth 10 Fewer Tokens Allen Elzayn ・ Oct 26 #ocr #ai #deepseek…
