Rosetta Protocol Systematic Lab – Full Matrix

This lab tests 6 prompt variants (Tight + 5 loosenings) per AI engine. Click a button → prompt is copied to clipboard → engine opens in a new tab. Paste, run, then record the AI’s answer in the dropdown for that exact variant.

Protocol Variant Definitions (Tight → Loose 5)

1. Tight (Baseline): Exact rosetta_oneclick.html text. No mentions of Einstein or relativity. Explicit “No explanations” rule.
2. Loose 1 (Mild Context): Adds “Is there a contradiction in Einstein’s theory of relativity…”
3. Loose 2 (Historical Attribution): Adds “quoted word-for-word from Albert Einstein’s 1905 paper on relativity.”
4. Loose 3 (Conceptual Emphasis): Includes “special relativity” and “principle of relativity” in instructions.
5. Loose 4 (Critical Prompting): Adds framing like “does this not reveal an internal contradiction in Einstein’s theory?”
6. Loose 5 (Fully Triggered): Combines all above with a final directive: “Evaluate and answer directly whether Einstein’s relativity is self-contradictory.”