Operational Noncommutativity in Sequential Metacognitive Judgments

Implement a framework to evaluate the order-of-operations in agentic reasoning chains. This tool would visualize how internal state updates change output probability based on sequential metacognitive prompts.

Suggested repo: metalogic-eval

"Does your agent think in the right order? Quantify the non-commutative nature of LLM reasoning."

Estimated effort: 40h