AI models have more than they show
A work-in-progress note on model interpretability, inner representations, and why next-token prediction may be too narrow an output paradigm.
These are my raw thoughts on building software, AI, and life. Feel free to reach out if any of these spark a conversation.
A work-in-progress note on model interpretability, inner representations, and why next-token prediction may be too narrow an output paradigm.
I wrote this on my last flight out of the city that broke me on day one and pieced together a better version of me over the three years I spent there.
A journey through building trust with Claude Code—from initiation to multi-agent workflows.
A review of one of the most unexpected years of my life and how it left me better off than I might otherwise have been.