Papers

• Theoretical Risks and Tabular Asterisks Paul Meehl
• Telling More Than We Can Know Nisbett & Wilson
• Adding Error Bars to Evals Evan Miller
• Deep Reinforcement Learning at the Edge of the Statistical Precipice Agarwal et al.
• Simple Random Search Provides a Competitive Approach to Reinforcement Learning Mania, Guy & Recht
• The 37 Implementation Details of Proximal Policy Optimization Huang et al.
• Revisiting the Arcade Learning Environment Machado et al.
• Categorizing Variants of Goodhart’s Law Manheim & Garrabrant
• Can a Biologist Fix a Radio? Yuri Lazebnik
• Could a Neuroscientist Understand a Microprocessor? Jonas & Kording
• Zoom In: An Introduction to Circuits Olah et al.
• A Mathematical Framework for Transformer Circuits Elhage et al.
• Interpreting GPT: The Logit Lens nostalgebraist
• Progress Measures for Grokking via Mechanistic Interpretability Nanda et al.
• SolidGoldMagikarp Rumbelow & Watkins
• Consciousness in Artificial Intelligence Butlin, Long et al.
• What Every Computer Scientist Should Know About Floating-Point Arithmetic David Goldberg
• Making Deep Learning Go Brrrr From First Principles Horace He
• Roofline Williams, Waterman & Patterson
• Anatomy of High-Performance Matrix Multiplication Goto & van de Geijn