🥶 Say goodbye to freezing hands. Water and windproof. Great for use with touch screen devices. Get it now: finesseattire.co/winter-gloves

Video lectures, Alberta UofA CMPUT 267 Machine Learning I fall 2025, by Vlad Tkachuk vladtkachuk4.github.io/machinelearnin… youtube.com/playlist?list=… .

EVolSplat4D: Efficient Volume-based Gaussian Splatting for 4D Urban Scene Synthesis Sheng Miao, Sijin Li, Pan Wang, Dongfeng Bai, Bingbing Liu, Yue Wang, Andreas Geiger, Yiyi Liao tl;dr: EVolSplat journal version->dynamic scene arxiv.org/abs/2601.15951

Stop predicting words one by one. A new paper just broke the next-token rule every LLM follows. It cuts generation steps by 4× and training compute by 44%. 𝗡𝗲𝘅𝘁-𝘁𝗼𝗸𝗲𝗻 𝗽𝗿𝗲𝗱𝗶𝗰𝘁𝗶𝗼𝗻 𝗶𝘀 𝘁𝗵𝗲 𝗯𝗼𝘁𝘁𝗹𝗲𝗻𝗲𝗰𝗸 CALM replaces word-by-word guessing with

Data Engineering Course with 20+ End-to-End Projects! What started as a few courses has now turned into a complete Data Engineering ecosystem. Today, our Data Engineering Course includes 20+ end-to-end projects and covers pretty much everything you need to crack interviews and

AI / ML Engineer in 2026, please learn: One ML stack deeply:- PyTorch or JAX, not just .fit(), but GPU memory, kernels, mixed precision, profiling, and why your model OOMs at 3am. Data:- Where it comes from, how it lies, how it drifts, how labels break, how leakage sneaks in,

Logical Intelligence introduces first energy-based reasoning AI Model, and brings Yann LeCun to leadership as founding chair of their Technical Research Board The 6-month-old Silicon Valley start-up, unveiled an “energy based” model called Kona and says it is more accurate and

FAR.AI is hiring engineers and scientists! Join one of our teams: - Deception - stress-test and train against lie detectors - Integrity - red-team frontier models and write evals for new threats - Foundations - develop infrastructure for scalable ML workflows NEW

A minimal Python CLI that helps you find and book plane tickets using tool calling and reasoning to solve multi-step workflows. Enjoy ↓ github.com/Liquid4All/coo…

LLM-in-Sandbox Elicits General Agentic Intelligence huggingface.co/papers/2601.16…

I would start with Video lectures, Stanford EE364A Convex Optimization I autumn 2023, by Stephen Boyd web.stanford.edu/class/ee364a/i… youtube.com/playlist?list=… .

Classic text on Outlier Detection in Temporal Data: amzn.to/2NZ64cQ ... including financial, business, medical, credit, sensor network (IoT / IIoT) time series data streams, plus spatiotemporal data, network data, ...

If you are interested in building an AI Scientist, I strongly encourage you to apply to Google DeepMind. While vibe coding and papering a demo is easy, creating a true AI Scientist requires decades of rigorous science and engineering and a large team of great culture—it is a

want to run a model on your phone? jalexine.github.io/performatively…

An open-source memory layer for AI agents – MemOS It gives agents a real memory operating system, where knowledge is structured, editable, and evolves over time, not disappearing into a black-box embedding store. MemOS enables: - Unified Memory API for adding, retrieving,

HOLY SHIT… GAMMA JUST SHOCKED THE $200B DESIGN INDUSTRY WITH ONE FREE TOOL. They just dropped Nano Banana Pro for FREE and it does what took me 6 hours in 60 seconds. I got this 25-slide Stardew Valley style PPT in just one shot and the results blew my mind.

Move over Buffon. Step aside Feller. Romanovsky flipped the coin 🪙 80,640(!) times — because Markov taught him not to trust words. Long before Monte Carlo simulations and “it converges eventually” hand-waving, probability was tested the hard way: by actually throwing the coin.

x.com/rohanpaul_ai/s…