01:24:35
ARC-AGI-3 Explained by the Team That’s Winning It
Machine Learning Street Talk convenes several members of a top-performing ARC-AGI-3 competition team — Jon Kotar, Stephano, D. Smith,...
Sort by: Newest Items
There are 95 items in this page
16:35
Introducing Ornith 1.0
Sam Witteveen introduces Ornith 1.0, a new family of open-weight models from Deep Reinforce that takes a fundamentally different appr...
08:51
OpenJarvis + Ollama: Local AI Agent That Tracks Every Watt
Fahd Mirza walks through the installation and hands-on testing of Open Jarvis, a newly released local-first personal AI framework dev...
12:18
Unlimited OCR from Baidu: One-shot Long-horizon Parsing: Run Locally
Baidu's Unlimited OCR introduces a fundamentally different approach to document parsing: processing entire multi-page documents in a...
09:36
Qwen3.6 (REAP 90pct GGUF): The Brain-Damaged Model
Fahd Mirza takes a deep look at an aggressively pruned variant of Qwen 3.6 — a 35-billion-parameter mixture-of-experts model — compre...
18:17
VibeThinker 3B – Taking on Giant Models
Sam Witteveen digs into VibeThinker 3B, a small language model from Waybo AI Lab — the AI research arm of the Chinese social network...
09:42
Luce Spark: Run a 35B Model Under 16GB VRAM Locally
Fahd Mirza demonstrates LuceSpark, a memory management technique that allows a 35-billion-parameter mixture-of-experts model to run w...
09:03
Gemma 4 Was Broken for Agents – Google Just Fixed It
Google's Gemma 4 12B model contained a subtle but impactful bug in its official Jinja chat template that was silently breaking multi-...
13:08
Gemma 4 12B QAT + MTP on llama.cpp Locally – Twice the Speed, Same Quality?
This video by Fahd Mirza walks through running Google's newly released Gemma 4 12B QAT (Quantization-Aware Training) model alongside...
06:49
Nanowhale-100m: Fascinating Implemention of DeepSeek-V4 Architecture
Fahd Mirza walks through Nanowhale-100M, a 110 million parameter language model built entirely from scratch—no borrowed weights—that...