PhDing in AI safety, currently poking at computational sparsity in LLMs
Current theme: default
Less Wrong (text)
Less Wrong (link)