American physicist and AI researcher at Anthropic who co-authored the Neural Scaling Laws paper — demonstrating that AI model performance improves predictably as a power law function of model size, data, and compute — one of the most consequential empirical findings in modern AI.
Discovered neural scaling laws — Kaplan et al.'s 2020 paper showed that language model performance scales predictably with model size, dataset size, and compute following smooth power laws, enabling researchers to predict model capabilities in advance and providing the theoretical foundation for the current 'scale up' approach to AI.
Discussing
Ask anything about Jared Kaplan — their ideas, influences, contradictions, legacy.
Sign in to start a conversation.