1 link tagged with all of: residual stream + activation inversion + maem + interpretability + sparse autoencoders
Click any tag below to further narrow down your results
Links
This arXiv paper proposes MAEMs, language models trained to turn a residual stream direction in another model into text that strongly activates it, rather than searching a fixed corpus for such text. The training signal is simply maximizing cosine similarity with the target direction, which lets the approach scale across many kinds of directions.
- A Qwen3.6-27B MAEM inverts held-out activations better than corpus search or a natural-language autoencoder, and beats corpus search at 10M tokens on most held-out languages and domains.
- It activates 89% of held-out SAE features and identifies concepts behind steering vectors roughly as well as NLAs.
- Because MAEM exemplars can be checked with a forward pass on the subject model, the method is more grounded than explanations from other metamodels.