A platform for structuring AI research and studying physics of AI
[ICML 2026] Learning Dynamics of Zeroth-Order Optimization: A Kernel Perspective
Learning as optimization & dynamical systems, autodiff, backprop, gradient flow from scratch.
Code release for "Learning Dynamics of Logits Debiasing for Long-Tailed Semi-Supervised Learning" (ICLR 2026)
Repository containing experimental pipelines for the paper "Learning Dynamics and Core of Learners".
Small toolkit to run teacher-student experiments and study learning dynamics with stochastic gradient descent (SGD).
Evidence-bounded manuscript and reproducibility artifacts for two synthetic studies of operational marker alignment and residual-associated marker separation in a discrete bottleneck.
Reproduction code for: Learning Rate Dynamics Across Training Phases in Mouse Visual Decision-Making (Princeton Senior Thesis, 2026)
To associate your repository with the learning-dynamics topic, visit your repo's landing page and select "manage topics."