ConceptAI & ML
Gradient ascent
An optimization technique that updates parameters by moving in the direction of the gradient to maximize a function.
Know first
Named as a prerequisite by the lectures that define this. The number is how many say so. A grey one has no page yet.
Opens up
Concepts that name this one as a prerequisite.
Lecturers
2Each in their own terms, from the easiest level.
Pieter Abbeel
An optimization procedure where you perform an update in the uphill direction for each coordinate based on the derivatives.
Lectures that assume it
8- Policy Gradients I: Deriving the REINFORCE Estimator & Policy Objective
- Deep Reinforcement Learning
- Policy Gradient Methods
- Deep Reinforcement Learning
- Machine Learning: Neural Networks and Decision Trees
- Language Models: Prompting, Chain-of-Thought, and Instruction-Tuning
- Visualizing and Understanding
- Aligning Open Language Models
Connections
20The lectures that define it or assume it, and the concepts next to it.