About Research

Using Claude to Accelerate Robot Dog Control in Robotics Mechanistic Interpretability of Large Language Models Clio: Analyzing Real-World Claude Usage While Preserving Privacy Character Training for Claude's AI Personality How AI Interpretability Reveals Claude's Internal Reasoning Model Welfare and the Moral Status of AI Language Models Circuit Tracing to Reveal Planning in Large Language Models Prompt Engineering Techniques for Large Language Models Reward Hacking and Emergent Misalignment in Large Language Model Training How Functional Emotions Emerge in Claude Through AI Interpretability The Societal and Economic Impacts of AI Systems Translating an AI Model's Internal Activations into Plain Language The Difficulty of AI Alignment Research in Large Language Models Sparse Autoencoders for Interpreting Neural Network Activations Emotional Support Use Cases of AI Chatbots AI Control as a Strategy for Mitigating AI Misalignment Project Vend: Claude Autonomously Running a Vending Machine Business Constitutional Classifiers for Blocking Universal Jailbreaks in Large Language Models Alignment Faking in Large Language Models Mechanistic Interpretability in Neural Networks AI-Enabled Cybercrime and Threat Intelligence Defense in Cybersecurity Model Welfare and the Possibility of Consciousness in AI Language Models