Turn-Averaged SAEs for Feature Discovery and Long-Context Attribution
Sparse autoencoders (SAEs) have become a useful tool for extracting interpretable features in language models. However, standard SAE architectures operate on individual token activations, meaning that the number of active features scales linearly with context length, and studying long model transcripts becomes difficult. We introduce turn-averaged SAEs, which represent a single Human or Assistant turn with a fixed number of features by learning to reconstruct the average model activation across the turn. We find that turn-averaged features describe a single turn's high-level characteristics more completely than per-token features when judged by an LLM. We also demonstrate that turn-averaged SAEs greatly simplify common downstream uses of SAEs like attribution graphs. Broadly, turn-averaged SAEs make interpretability techniques practical at long context lengths.
Code (0)
등록된 구현이 없습니다.
Similar Papers 제목 키워드 기반
Llama Scope: Extracting Millions of Features from Llama-3.1-8B with Sparse Autoencoders
Sparse Autoencoders (SAEs) have emerged as a powerful unsupervised method for extracting sparse representations from language models, yet scalable training remains a significant challenge. We introduce a suite of 256 SAE…
Language ModelingLanguage ModellingMultimodal Model Diffing for Feature Discovery and Control
Multimodal Large Language Models (MLLMs) exhibit strong visual understanding, yet the internal features that cause these behaviors remain difficult to identify, audit, or control. While applicable to post-hoc inspection,…
Improving Dictionary Learning with Gated Sparse Autoencoders
Recent work has found that sparse autoencoders (SAEs) are an effective technique for unsupervised discovery of interpretable features in language models' (LMs) activations, by finding sparse, linear reconstructions of LM…
Dictionary LearningFeatures that Make a Difference: Leveraging Gradients for Improved Dictionary Learning
Sparse Autoencoders (SAEs) are a promising approach for extracting neural network representations by learning a sparse and overcomplete decomposition of the network's internal activations. However, SAEs are traditionally…
Dictionary LearningWhich Sparse Autoencoder Features Are Real? Model-X Knockoffs for False Discovery Rate Control
Although sparse autoencoders (SAEs) are crucial for identifying interpretable features in neural networks, it is still challenging to distinguish between real computational patterns and erroneous correlations. We introdu…