Activation steering
Activation steering is a family of inference-time techniques in mechanistic interpretability and AI safety that modify a neural network's internal activations to influence its behavior, without retraining the…
Interpretability
Explore Interpretability through related topics and the articles other pages reference most.
Explore articles
Browse subtopics: Large Language Models
Articles that also belong to these categories. Counts cover all of Interpretability.
All articles (44)Machine Learning (12)Anthropic (8)Deep Learning (8)AI Safety (6)Neural Networks (6)AI Research (4)Large Language Models (4)Computer Vision (2)Developer Tools (2)Open Source AI (2)Transformer Models (2)AI Companies (1)AI Ethics (1)Algorithms (1)Generative AI (1)Mathematics (1)Model Evaluation (1)OpenAI (1)People (1)
Showing 1-4 of 4 articles
Patchscopes
Patchscopes is an interpretability framework for inspecting hidden representations of large language models by patching an internal activation from a source computation into a separate target inference whose…
Persona vectors
Persona vectors are single linear directions in the activation space of a large language model that correspond to high level character traits such as evil, sycophancy, or a propensity to hallucinate.
Refusal direction
The refusal direction is a finding from mechanistic interpretability research that the refusal behavior of safety fine-tuned chat language models is mediated by a single