Applying sparse autoencoders to unlearn knowledge in language models

We investigate whether sparse autoencoders (SAEs) can be used to remove knowledge from language models. We use the biology subset of the Weapons of Mass Destruction Proxy dataset and test on the gemma-2b-it and gemma-2-2b-it language models. We demonstrate that individual interpretable biology-related SAE features can be used to unlearn biology-related knowledge with minimal side-effects. Our results suggest that negative scaling of feature activations is necessary and that zero ablating features is ineffective. We find that intervening using multiple SAE features simultaneously can unlearn multiple different topics, but with similar or larger unwanted side-effects than the existing Representation Misdirection for Unlearning technique. Current SAE quality or intervention techniques would need to improve to make SAE-based unlearning comparable to the existing fine-tuning based techniques.

An Adversarial Example for Direct Logit Attribution: Memory Management in gelu-4l

We provide concrete evidence for memory management in a 4-layer transformer. Specifically, we identify clean-up behavior, in which model components consistently remove the output of preceeding components during a forward pass. Our findings suggest that the interpretability technique Direct Logit Attribution provides misleading results. We show explicit examples where this technique is inaccurate, as it does not account for clean-up behavior.