Chris Olah - Looking Inside Neural Networks with Mechanistic Interpretability скачать в хорошем качестве

Chris Olah - Looking Inside Neural Networks with Mechanistic Interpretability

"Looking Inside Neural Networks with Mechanistic Interpretability" by Chris Olah. Delivered at the 2023 San Francisco Alignment Workshop. Image Credits Glenn Carstens-Peters Markus Spiske Alexander Sinn Brett Jordan National Cancer Institute Alex Kalligas Clay Banks Nick Fewings Patrick Hendry Papers Carter et al. (2019). Exploring Neural Networks with Activation Atlases. Distill. Olah et al. (2020). Zoom In: An Introduction to Circuits. Distill Circuits Thread. Olah et al. (2020). Naturally Occurring Equivariance in Neural Networks. Distill Circuits Thread. Olsson et al. (2022). In-context Learning and Induction Heads. Transformer Circuits Thread. Elhage et al. (2022). Toy Models of Superposition. Transformer Circuits Thread.