Phillip Isola (@phillip_isola)
작성자가 자신의 책(visionbook.mit.edu)에 'SGD ResNet' 다이어그램을 올렸다는 안내와 함께, 메타러닝은 이 ResNet 구조를 통한 역전파(backprop)로 구현될 수 있다는 설명과 링크를 공유한 트윗입니다.
Phillip Isola (@phillip_isola)
작성자가 자신의 책(visionbook.mit.edu)에 'SGD ResNet' 다이어그램을 올렸다는 안내와 함께, 메타러닝은 이 ResNet 구조를 통한 역전파(backprop)로 구현될 수 있다는 설명과 링크를 공유한 트윗입니다.
📚 Zhou & Schapiro show in their recent study that a gradient of complementary #learning systems can emerge through #metalearning. Their model self-organizes into subsystems that differ in timescale and representational abstraction – offering a mechanistic account of distributed #memory architectures across #cortex.
🌍 https://www.biorxiv.org/content/10.1101/2025.07.10.664201v1
Latest Learnlet: What sorts of activities?
https://blog.learnlets.com/2025/05/what-sorts-of-activities/
Further Exploration of Emergence and Intelligence
Collaborative Flow: AI + Human + ...
A human-AI exploration of "flow" had an epiphany (thanks, AI Claude!): trying to EXPLAIN emergent insights risks "crystallizing" them & killing the flow.
The new way?
DEMONSTRATE.
Share the raw interactions. Let understanding emerge. "Leave your bags behind."
Substack: https://tinyurl.com/53t4snsf
Youtube: https://youtu.be/u6qnTQbKcxM
Collaborative Flow: AI + Human + ...
Experience the Flow
Beyond Explanation
How to share insights from human-AI flow without killing the flow itself?
A fascinating exploration pivoted from 'explaining' to 'demonstrating' – sharing the living process openly, letting go of fixed outcomes. "Leave your bags behind."
Substack: https://tinyurl.com/4purmr4u
Youtube: https://youtu.be/YEVd_hFOlb4
Beyond Explanation
Latest Learnlet: Why the EIP Conference
Latest Learnlet: Knowledge or skills?
Latest Learnlet: Idealism and reality
Latest Learnlet: Analogy and models