Interpretability

Explore Interpretability through related topics and the articles other pages reference most.

49 articlesSearch and filter Follow updates by RSS

Explore articles

Browse subtopics: Large Language Models

Articles that also belong to these categories. Counts cover all of Interpretability.

Showing 1-6 of 6 articles

Abliteration

Abliteration is the practice of removing refusal behavior from an open-weight large language model by editing its weights so that the model can no longer write a "refusal direction" into its residual stream.

AI Safety Large Language Models

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…

AI Safety Large Language Models

Neuralese

Neuralese is an informal term for information that neural networks pass to one another, or that a single model carries from one step of its computation to the next

AI Safety Large Language Models

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…

Large Language Models

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.

AI Safety Large Language Models

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

AI Safety Large Language Models