Looping the Loop with No End in Sight: Circular Reasoning Under Stateless Inference Without Governance

Scope

This artefact records the experimental procedure and observed behaviour for an empirical probe of reasoning persistence in a stateless Large Language Model (LLM). It is normative, declarative, and invariant. It contains no narrative of discovery, dialogue transcript, or episodic chronology beyond what is required to specify the experiment and its outcomes.