A probabilistic causal-graph physician: exact noisy-OR inference over a causal knowledge graph, with decision-theoretic stopping that picks the most informative next question.
Doc is a probabilistic causal-graph physician, fronted by Macleod the cat and powered by an exact inference engine.
How it reasons
Causal knowledge graph. Findings link to diseases through a causal graph built from public sources.
Exact noisy-OR inference. Every answer updates the probability of every hypothesis exactly. There is no sampling and no LLM guesswork in the maths.
Decision-theoretic stopping. Doc picks the next question by its expected value of information, and stops when another question would not change the decision.
The diagnosis views show the live causal graph, the ranked differential and how each answer moved it.
