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.
Interpretability
Explore Interpretability through related topics and the articles other pages reference most.
Explore articles
Browse subtopics: AI Safety
Articles that also belong to these categories. Counts cover all of Interpretability.
All articles (49)Machine Learning (12)AI Safety (9)Anthropic (9)Deep Learning (8)Neural Networks (8)AI Research (6)Large Language Models (6)Open Source AI (3)Computer Vision (2)Developer Tools (2)People (2)Transformer Models (2)AI Agents (1)AI Ethics (1)AI for Science (1)Algorithms (1)Generative AI (1)Mathematics (1)Model Evaluation (1)Natural Language Processing (1)OpenAI (1)Training & Optimization (1)
Showing 1-9 of 9 articles
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…
Mechanistic interpretability
Mechanistic interpretability (often abbreviated as mech interp or MI) is the field that reverse-engineers the internal computations of neural networks, particularly transformers
Neural Networks Are Already Clean
Neural Networks Are Already Clean is a long interactive research essay by Nick Cammarata, published by the interpretability company Goodfire on 15 October 2026 at already-clean.goodfire.com.
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
Nick Cammarata
Nick Cammarata is a machine learning researcher who works on mechanistic interpretability, the study of how the internal components of neural networks compute.
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
Representation Engineering
Representation Engineering (often abbreviated RepE) is a top-down approach to artificial-intelligence transparency and control that reads and manipulates high-level concepts (such as honesty, harmlessness, and…