flâneur — a map of the web's best reading

Leela Interpretability

leela-interp.github.io · saved by 1 readers

One of our early experiments was to do activation patching. We patch a small part of Leela's activations from the forward pass of a corrupted version of a puzzle into the forward pass on the original puzzle board state. Measuring the effect on the final output tells us how important that part of Leela's activations was.

Explore this link on the map →

related reading