Kerem Zengin https://kerem-z.github.io/ Essays and notes on type theory, explainable AI, category theory, cognitive science, and the philosophy of mind. Teaching assistant courses at METU IAM. en-us Wed, 12 Aug 2026 12:00:00 GMT Types as constraints https://kerem-z.github.io/blog/types-as-constraints-demo/ https://kerem-z.github.io/blog/types-as-constraints-demo/ Wed, 12 Aug 2026 12:00:00 GMT A playable sketch of type refinement: drag the precision dial and watch what gets forbidden. Formal Methods in ML Interpretability https://kerem-z.github.io/blog/formal-interpretability/ https://kerem-z.github.io/blog/formal-interpretability/ Mon, 03 Aug 2026 12:00:00 GMT What would it mean for an explanation to be correct, not merely persuasive? What is an explanation? https://kerem-z.github.io/pieces/what-is-an-explanation/ https://kerem-z.github.io/pieces/what-is-an-explanation/ Thu, 09 Jul 2026 12:00:00 GMT Climb an explanation ladder: saliency, counterfactuals, causal claims, formal guarantees. Causal Structures in Neural Networks https://kerem-z.github.io/blog/causal-structures/ https://kerem-z.github.io/blog/causal-structures/ Thu, 02 Jul 2026 12:00:00 GMT Do-calculus as an epistemic tool for reading model decisions, not a dashboard feature.