ImageNet
includes ImageNet 22K
1 statements across 1 episodes · 0 bullish · 0 bearish · 1 people on the record · first statement Feb 12, 2026 by Jeff Dean · said 34 times in 16 episodes since 2023 · across every show →
Mentions by year, the whole family
brought up most by Ari Morcos (6), Michael Royzen (3), Jeff Dean (3), Justin Johnson (2), Jeremy Howard (2), Fei-Fei Li (2), Shawn Wang (1), RJ Honicky (1)
tap a year for its mentions
2026 6 mentions in 3 episodes 2 per episode
- The AI Frontier: from Gemini 3 Deep Think distilling to Flash — Jeff Dean
- 🔬 Training Transformers to solve 95% failure rate of Cancer Trials — Ron Alfa & Daniel Bear, Noetik
- Simulating Humanity: from Generative Agents to 8 Billion Digital Twins — Joon Sung Park, Simile AI
- every mention in 2026, scene by scene →
2025 17 mentions in 6 episodes 3 per episode
- After LLMs: Spatial Intelligence and World Models — Fei-Fei Li & Justin Johnson, World Labs
- Better Data is All You Need — Ari Morcos, Datology
- SAM 3: The Eyes for AI — Nikhila & Pengchuan (Meta Superintelligence), ft. Joseph Nelson (Roboflow)
- Why RL Won — Kyle Corbitt, OpenPipe (acq. CoreWeave)
- Information Theory for Language Models: Jack Morris
- The Magic of LLM Distillation — Rishabh Agarwal, Google DeepMind
- every mention in 2025, scene by scene →
2024 6 mentions in 5 episodes 1 per episode
- Breaking down the OG GPT Paper by Alec Radford
- Best of 2024 in Vision [LS Live @ NeurIPS]
- [Paper Club] Intro to Diffusion Models and OpenAI sCM: Simple, Stable, Scalable Consistency Models
- In the Arena: How LMSys changed LLM Benchmarking Forever
- llm.c's Origin and the Future of LLM Compilers - Andrej Karpathy at CUDA MODE
- every mention in 2024, scene by scene →
2023 5 mentions in 2 episodes 3 per episode
Everything said about ImageNet, oldest first
Feb 12, 2026
Jeff Dean: Distillation originated to compress 50-model ensembles into serviceable form
“Distillation was originally motivated because we were seeing that we had a very large image data set at the time, you know, three hundred million images that we could train on with, you know, I forget, like 20,000 categories or something, so much bigger than I…”