Technical writing on model training, agents, observability, and agent architecture. https://dunkeln.github.io/ en-us Better tooling happens when MCPs balance the cognitive load for client agents. https://dunkeln.github.io/writing/supercharging-a-service-with-mcp/ https://dunkeln.github.io/writing/supercharging-a-service-with-mcp/ Fri, 31 Jul 2026 00:00:00 GMT Mobisim turns a raw GLB vehicle into something you can talk to using structural DAG extraction, a reviewed semantic overlay, real-time WebRTC, and deterministic execution. https://dunkeln.github.io/writing/grounded-voice-copilot/ https://dunkeln.github.io/writing/grounded-voice-copilot/ Tue, 05 May 2026 00:00:00 GMT Designing an RLVR-style verifier and scalar reward function for agent evaluation workflows, with judge metrics, render-validity gating, and reward diagnostics for stability, win rate, and failure modes. https://dunkeln.github.io/writing/rl-verifier-design/ https://dunkeln.github.io/writing/rl-verifier-design/ Sat, 21 Feb 2026 00:00:00 GMT Research notes on modeling stochastic cellular automata with entropy-based patching and transformer architectures for next-state prediction. https://dunkeln.github.io/writing/cellular-automata-prediction/ https://dunkeln.github.io/writing/cellular-automata-prediction/ Mon, 09 Jun 2025 00:00:00 GMT