🧠 Mixture of Cognitive Reasoner (MiCRo) Expert Routing Visualizer

Enter a prompt (and optionally an assistant reply), pick a model, and visualize how tokens were routed across experts.

This demo visualizes how modular language models allocate computation across specialized experts—Language, Logic, Social, and World—when processing a given prompt. Each expert corresponds to a cognitive domain inspired by human brain networks. Enter a prompt to see how tokens are dynamically routed across modules, revealing the model's internal reasoning structure.
Select a model

Ablate Experts

(Check to disable an expert; the routing percentages will be redistributed among the remaining experts)

Routing Percentages

Try these examples:
Select a model User prompt Assistant prompt (optional) Language Expert Logic Expert Social Expert World Expert