Meridian Cambridge
Meridian is a hub for people working on the long-term future of humanity. Based in Cambridge, UK, we support work on AI safety, biosecurity, and other global challenges through events, fellowships, community-building, and research. Subscribe for talks, interviews, and project highlights from the Meridian community.
Lecture 16 • The Future of Language Models
Lecture 15 • AI Orgs and Agendas
Lecture 14 • AI Control and Safety Cases
Lecture 13 • Model Evaluations
Lecture 12 • Risks from Advanced AI
Lecture 11 • Deceptive Alignment and Alignment Faking
Lecture 10 • Out-of-Context Reasoning and Situational Awareness
Lecture 09 • Reward Hacking and Goal Misgeneralisation
Lecture 08 • Optimisation and Reasoning
Lecture 07 • Agents and Agent Architectures
Lecture 06 • Reward Modelling
Lecture 05 • Reinforcement Learning for Language Models
Lecture 04 • Post-Training Language Models
Lecture 03 • Neural Scaling Laws
Lecture 02 • The Transformer Architecture
Lecture 01 • Introduction to Language Models