profile pic
⌘ '
raccourcis clavier
Search

Automatic Differentiation

  • Étiquette

    • publié à

    • modifié à

    • durée

      1 min de lecture

    • source

      llms.txt

    see also: Autograd and Jax

    Input: code compute a function Output: code compute the derivative of the function

    AD writes functions as sequence of compositions block f(x)=fn∘fn−1∘…∘f1(x)f(x) = f_n \circ f_{n-1} \circ \ldots \circ f_1(x), and then computes the derivative of the function by applying the chain rule.

    Vous pourriez aimer ce qui suit

  • Real-time system a la carte
  • Bias and intercept
  • game theory
  • scripts.
  • malleable software
  • gradient descent
  • Orwellian
    • Machine learning
    Detects pattern within data and use it to make useful prediction. Generally AI \subset ML \subset DL Some main exploration: Transformers Large language models NLP CNN Logistic regression Optimization gradient descent hyperparameter tuning ensemble learning Recommender systems Reinforcement learning Q-learning Policy Gradient Monte-Carlo Tree Search Generative Models GAN VAE Autoencoder sparse autoencoder sparse crosscoders Supervised Q-learning Low-rank adapters Fields mechanistic interpretability Related: linear algebra.
    Vector calculus
    just enough vector calculus to be dangerous