Yash Naik·Oct 1𝗧𝗵𝗲 𝗡𝗘𝗦𝗧 𝗟𝗮𝗯 𝗡𝗼𝘁𝗲𝘀 — 𝗟𝗼𝗴 𝟬𝟭. 𝗡𝗼𝘁𝗲 𝟬𝟯Sixteen gigabytes. That’s the whole budget, and no amount of wanting more changes the number on the box. Eleven minds, a 16GB RTX 5070 Ti…
Yash Naik·Sep 25The NEST Lab NotesEleven minds. One block. That’s the whole bet, stated as plainly as I can state anything at this hour. Not eleven architectures tuned by…
Yash Naik·Sep 18The NEST Lab NotesThe first thing I built wasn’t a neuron. It was a court system. A machine that can’t lie to itself, can’t quietly forget what it did, and…
Yash Naik·Sep 17Building eleven minds on one GPU, alone, on purpose: The NEST Lab NotesI have officially crossed my thinking capabilities. My GPU is sitting at 78°C like it’s personally offended by what I’m asking of it…
NYU Center for Data Science·Mar 13Generating the Stepping Stones: A New Framework for AI Self-ImprovementWhen training AI models, the most difficult and specialized tasks are often precisely the ones with the least available data. To overcome…
Juan Pablo Ávila Páez·Jan 11Enhancing Proximal Policy Optimization (PPO) with Curriculum LearningDisclaimer: I acknowledge the use of Claude 4.5 Opus/Sonnet (Anthropic) and Gemini 3 Pro/Flash (Google) for brainstorming ideas, aiding…A response icon2A response icon2
InAI-ML Interview PlaybookbySajid Khan·Nov 26, 2025What is Curriculum Learning in Machine LearningHow Models Learn Better by Starting Simple
Bhagya Rana·Sep 25, 2025Curriculum Learning, Real GainsPace your data so models learn faster — and forget less
Mohamed Aymen BOUYAHIA·Sep 6, 2025From Easy to Hard: Understanding Curriculum Learning in AIWhen training machine learning models, the standard approach is to feed data in a random order. However, this isn’t how humans learn !
NYU Center for Data Science·Jul 1, 2025Neural Networks Need Kindergarten: Training AI Like Animals LearnNeural networks achieve rat-like performance in long-term decision tasks when they’re pretrained on fundamental cognitive skills.
