What Does 35B-A3B Mean? Mixture of Experts Explained
Those "35B-A3B" names on every new open model, decoded: what Mixture of Experts (MoE) is, why a 35-billion-parameter model only uses 3 billion at a time, why that makes AI cheaper, and how it fits on a phone.
Up next
All videos →
What Are Tokens in AI? Why ChatGPT Couldn't Spell Strawberry (Explained Simply)
What Do Q4 and Q8 Mean? I Squeezed One AI Into 9 Versions
What Is AI Distillation? Your DeepSeek R1 Isn't Really R1
Is AI Really a Black Box? I Opened Up NVIDIA's Nemotron
Claude Opus 5.5 vs GPT-6 Astra: Who Can Beat the Original Celeste First?
GPT-6.1 Terra Just Dropped. I Actually Tested It.
Opus 5.5 Motion Graphics: Zero After Effects, All Code
OpenAI and Claude Just Disproved a Famous Math Conjecture. Is P = NP?





