This audit walks the full cipher version chain, scoring each version on the same 107-element benchmark and documenting *what changed* and *what was learned* at each step. The purpose: anyone evaluating the cipher's predictive performance needs to understand which version is the canonical one, why previous versions stand or fall, and what the data actually says about which mechanisms were load-bearing.
\n| Version | \nDate | \nBench score (107 elements) | \nMedian angular err | \nKey mechanism | \nStatus | \n
|---|---|---|---|---|---|
| v7 | \n2026-03-22 | \n96.9% property match | \n\u2014 | \nDimensional architecture (2D/3D/4D regimes) | \nSuperseded | \n
| v8 | \n2026-03-23 | \n100% on ductility / conductor / band gap | \n\u2014 | \nLattice resonance at coord/{2,3} | \nSuperseded | \n
| v9 | \n2026-03-31 | \n71.9% exact / 94.8% A+P | \n3.2\u00b0 | \nEigenvalue susceptibility + snap/tunneling | \nSuperseded | \n
| v10 | \n(experimental, not released) | \nregression | \n\u2014 | \nSpherical-limit correction layer | \nDiscarded | \n
| v11 | \n2026-04-05 | \n71/107 corrected / 98.1% A+P | \n1.0\u00b0 | \nBicone + snap + Thomson/Fibonacci hybrid | \nSuperseded | \n
| v12 | \n2026-04-30 | \n82/107 (perspective 91/107, 0 MISS, 100% A+P) | \nclean | \nPure terrain, no corrections | \nCanonical | \n
F1. v9 \u2192 v11 was honest progress. Each version added a specific mechanism (eigenvalue susceptibility, snap-state tunneling, bicone simultaneous derivation) and each improved the score on the same bench. The 98.1% A+P on v11 was a genuine result.
\nF2. v10 was the warning shot. v10 tried to add a spherical-limit correction layer (the cipher's misses approach spherical limit; correct that and the misses should go away). It regressed to 20.9% exact \u2014 worse than v8. *Result:* uniform correction layers don't work. v10 was never released; v11 went back to per-element progressive blending with much better results.
\nF3. v12 falsified the corrections approach entirely. The v11 \u2192 v12 transition stripped *all* corrections (eigenvalue layer, snap layer, shell-completion logic) and returned to pure terrain. Result: 82/107 clean vs v11's 71/107 corrected on the same bench. Every correction layer was *hurting* accuracy. See the corrections-hurt-accuracy log for the head-to-head numbers.
\nF4. v12 perspective view: 91/107 with 0 outright misses. When v12 is read with a perspective lens (CN\u22646 \u2192 2D regime, CN\u22657 \u2192 3D regime), the score climbs to 91/107 with zero outright misses and 100% A+P. This is the framework's best-ever score on the bench.
\nF5. The c-ladder correction (Tribonacci replaces Fibonacci/8 for dims 4\u20136) is orthogonal to versions. It changed the c-values for d \u2265 4 but didn't affect cipher predictions for crystal structures (those live in cycle 1, d \u2264 3). See the c-ladder correction trail audit.
\nF6. v11's \"complete self-derivation\" claim stands. v11 was the version that proved the cipher *can* predict every element from Z alone with no external data. v12 demonstrates that the *terrain* carries the predictive power, but v11 was the milestone that closed the self-derivation chain. The v11 entry stays preserved as the milestone.
\nF7. Historical versions preserved, not deleted. Every version's results, code, and analysis remain published with status: superseded where appropriate. No silent rewriting. The audit trail is the intellectual-honesty record.
read 91/107).
\nforward pointer to v12.
\n(Paper 5 done, Papers 1\u20133 pending \u2014 see paper-status entries).
\nThe v9 \u2192 v10 \u2192 v11 \u2192 v12 trajectory taught the framework's no-corrections rule: every attempted correction layer reduced accuracy. The reason is that corrections impose a *human-set* fix on top of the geometric terrain; the terrain, when allowed to speak for itself, is more predictive than terrain-plus-correction. This rule now applies to all future cipher work and to the framework more broadly (see feedback_cipher_no_corrections \u2014 the rule is durable).
\nThe non-obvious part: this is the *opposite* of standard scientific practice, where models are improved by adding correction terms. Here, every addition *hurt*. The reason it works that way for the cipher is that the underlying mechanism is geometric, and the geometry is exact \u2014 not approximate. There's nothing to \"correct\" because nothing was approximated.
\nThe cipher framework has gone through versions v7 through v12 (with v10 discarded). This audit lays out the version chain on a single 107-element bench so anyone evaluating the cipher knows which version is canonical and what each version contributed.
\nCanonical version: v12. Pure terrain, no corrections, scores 82/107 raw and 91/107 with 0 outright misses under the perspective read (CN\u22646 \u2192 2D, CN\u22657 \u2192 3D). 100% A+P. Best-ever bench score.
\nv11 stays preserved as the milestone that closed the self-derivation chain \u2014 proved the cipher *can* predict every element from Z alone, 98.1% A+P, median 1.0\u00b0 angular error.
\nThe big finding from the trajectory: every attempted correction layer (v10 spherical limits, v11 eigenvalue blending, others) *hurt* accuracy on the same bench. v12's improvement came from *stripping* corrections, not adding them. This is the opposite of standard practice, and it taught the framework's no-corrections discipline.
\nWhy corrections hurt here: the underlying mechanism is geometric and exact, not approximate. There's nothing to correct because nothing was approximated. Adding a correction layer imposes a human-set fix on top of an already-exact terrain.
\nStatus: confirmed. All historical versions preserved, all superseded ones tagged. Papers that reference earlier versions get status notes (Paper 5 done, others pending). The audit trail is the record.
", "see_also": [ "cipher-v11-complete-self-derivation", "cipher-corrections-hurt-accuracy", "c-ladder-correction-trail" ], "cited_by": [ "cipher-v8-lattice-resonance", "external-ai-audit-protocol", "four-blind-tests-audit", "non-euclidean-measurement-discovery", "paper-1-status-2026-05", "paper-3-status-2026-05", "paper-revision-sync-audit", "triangle-regression-v12-foundation" ], "attachments": [], "schema_version": "1.0", "generated_at": "2026-05-12T03:27:18.533879Z" }