The Science
The measurements described in these pages are our own inventions, built on the published research and pushed past it where the research stopped. We hold the implementation details close — that is the nature of the edge — but we are working on ways to publish evidence of the system's quality that do not give the machinery away. Until then, the strongest evidence is the kind you generate yourself: run a session. It costs you nothing but the minutes.
The details, one topic per page:
- Shannon Entropy (ΔS)
- The 2000 result, its three faults in practice, and the measure we run instead.
- Mutually Distinct Targets
- Vision models, classical passes, and sets scored by their weakest link.
- Research-Derived Protocol Rules
- No backgrounds, one subject per image, and never the same target twice.
- Knowing When Not to View
- Geomagnetic conditions, watched for you before you begin.
- A Protocol Judge Available 24/7
- Independent SRI-scale judging in about a minute — and what the blindness research actually found.
