Person
Christopher Olah
Appears alongside
Featured in threads
Tracks
- Safety & alignment 3
- Ideas & essays 1
A Goodfire researcher argues neural networks are far more interpretable than the field believed
Nick Cammarata estimates about 90% of Qwen3-32B's neurons have one main job, and says the 2020 superposition evidence he co-authored was misread; a companion paper is still to come.
Safety & alignment · Ideas & essays
Anthropic publishes 'Towards Monosemanticity'
Sparse autoencoders decomposed a single 512-neuron layer into more than 4,000 human-interpretable features, far more than the raw neurons showed.
Safety & alignment
Anthropic publishes 'Toy Models of Superposition'
Elhage, Olah and colleagues showed small networks represent more features than they have neurons by packing them into overlapping directions, complicating efforts to read a model's internals.
Safety & alignment