1 case · sorted by most recent

Cases

Bounded investigations with a question, method, evidence, and a record of what changed. Each Case exposes its starting model and the evidence that survived the challenge.

• August 2026 · Experiment

  PLACEHOLDER — Does explicit sampling widen the useful search space of an LLM?

  Placeholder Case. Real content, evidence, and artefacts will land here after the first BN voice pass.

  AI & learning · Language & judgment