TopK SAE
No mentions found
This entity hasn't been tracked yet, or Iris is still building its knowledge base.
Related Articles from SNS
Toward Identifiable Sparse Autoencoders
Announce Type: new Abstract: Recently, sparse autoencoders (SAEs) have emerged as an attractive tool for interpreting and interacting with representations in practical neural networks. While it is common empirical folklore, we also show theoretically that SAEs are highly unstable: different training runs are likely to produce different concept dictionaries and sparse codes. We characterize the model properties that hinder the stability of real-world SAEs, and address each of these problems...
Pre-Intervention Prediction of Sparse Autoencoder Steering Side Effects
Announce Type: new Abstract: Sparse autoencoder (SAE) features are increasingly used to steer language models, but feature steering is rarely clean: the same intervention can behave inconsistently across contexts and perturb unrelated features. We introduce a pre-intervention screening framework for forecasting SAE steering side effects from feature statistics computed before steering. We operationalize side effects along two axes of steering modularity, effect stability and collateral...
Vision-Language Asymmetry in Bistable Image Captioning
arXiv:2606.08031v1 Announce Type: new Abstract: Wittgenstein's duck-rabbit poses a question for vision-language models: when a model captions an ambiguous image, where in the model is the commitment to one aspect made? We address this with a 3,320-generation behavioral baseline over 83 bistable stimuli that surfaces three regimes (default-dominant, force-dominant, force-balanced) under neutral vs forced-choice prompting, then probe the underlying representations using a TopK sparse...
On the Relationship Between Activation Outliers and Feature Death in Sparse Autoencoders
arXiv:2605.31518v1 Announce Type: new Abstract: Sparse autoencoders (SAEs) decompose neural network activations into interpretable features, but many learned features never activate, a problem called feature death that wastes dictionary capacity and can reintroduce superposition. Death rates vary dramatically between models: near-zero on GPT-2, over 70% on AlphaFold3 with identical configurations. We find that dimension-level activation outliers (dimensions whose mean magnitude is large...