# Exception: reverse-engineer \(12.48\,\mathrm{fm}\) — no non-circular identity

### \(p\,W_{\mathrm{Cl}}=6\pi^2\) is a \(0.20\%\) near-miss · not acceptance · \(3L_{\mathrm{knot}}\), \(\pi L_{\mathrm{knot}}\) stay FAIL · no promote

**CPT Framework — LSTAR-CAND / FB-O1′ / Clifford \(W=2\pi^2\)**  
**Date:** 2026-09-17  
**Status:** **EXCEPTION + PROBE** · `PASS_REVERSE_EXCEPTION_NO_NONCIRCULAR_IDENTITY` · user-requested reverse-engineering · FB-O1′ **stays CLOSED F** · no promote  
**Stimulus:** User — exception: reverse-engineer the \(12.48\,\mathrm{fm}\) parameter to rule every avenue out.  
**Parents:** `a_geom_p1_lstar_new_primitive_candidates.md` · `a_geom_p1_fb_o1_prime_overlap_class_results.md` · `a_geom_p1_willmore_knot_host_scale.md`  
**Probe:** `scripts/investigations/s3_geometry/p1_lstar_reverse_engineer_exception_probe.py`  
**Data:** `scripts/investigations/s3_geometry/data/p1_lstar_reverse_engineer_exception.json`

Tags: **[E honesty]** · **[E algebra]** · **[E process]** · **[Programme]**

---

## 0. One-line

> **The exception was run. Nothing non-circular equals \(12.48\,\mathrm{fm}\).** Identities are \(E_0 R_C\), \((\sqrt{3}/4)a_{0,N}\), and \(\hbar c/(2pq U_0)\) — the same algebra. The only interesting construct is banked Clifford Willmore: \(p\cdot 2\pi^2=6\pi^2\) vs \(E_0=\sqrt{3}/(4\alpha)\), relative \(2.0\times 10^{-3}\) (\(\Delta S=-0.12\)). That is **inside** the \(1.17\%\) \(\Phi\)-cut and **not** a proof. Pre-registered \(3L_{\mathrm{knot}}\) and \(\pi L_{\mathrm{knot}}\) stay FAIL. FB-O1′ stays closed. No promote.

---

## 1. Scope of the exception

Cartesian product of **banked lengths** (\(R_C\), \(R_p\), \(L_{\mathrm{knot}}\), \(a_{0,N}\), \(\hbar c/U_0\), …) with **named factors** (\(p,q,pq,\varepsilon,\sqrt{3},\pi,2\pi^2,1/\alpha,\ldots\)). Cut: report anything within \(5\%\) or \(|\Delta S|\le\ln 2\). **Accept only symbolic identity.** The \(1.17\%\) band disqualifies worse misses; it does **not** accept a near-miss.

---

## 2. What hits

| Construct | \(\ell\) (fm) | vs \(\ell_\star\) | Verdict |
|-----------|-------------:|-------------------|---------|
| \(E_0 R_C\) | 12.4796 | \(0\) | **CIRCULAR** |
| \((\sqrt{3}/4)(R_C/\alpha)\) | 12.4796 | \(0\) | **CIRCULAR** (nickname) |
| \(\hbar c/(2pq U_0)\) | 12.4796 | \(0\) | **CIRCULAR** |
| \(p\cdot W_{\mathrm{Cl}}\,R_C=6\pi^2 R_C\) | 12.4542 | \(-0.203\%\) | **NEAR_MISS_NOT_ACCEPT** |
| \(\pi L_{\mathrm{knot}}\) | 12.6924 | \(+1.71\%\) | **FAIL_NUM** (PF2) |
| \(p L_{\mathrm{knot}}\) | 12.1204 | \(-2.88\%\) | **FAIL_NUM** (PF1) |

Clifford \(W=2\pi^2\) is banked. \(p\times W_{\mathrm{Cl}}=6\pi^2\approx 59.218\) vs \(E_0\approx 59.338\). Minutes-class \(\Phi\) would still be \(O(1)\). It is **not** \(\sqrt{3}/(4\alpha)\). Accepting it would be the method the \(1.17\%\) rule forbids.

No other combo in the dictionary is an identity.

---

## 3. Integrity

| Gate | Verdict |
|------|---------|
| Exception run; dictionary scored | **PASS** |
| Non-circular symbolic \(\ell_\star\) | **NO** |
| Accept \(6\pi^2\) / unpark FB-O1′ | **NO** |
| Promote | **NOT APPROVED** |

**Exit:** `PASS_REVERSE_EXCEPTION_NO_NONCIRCULAR_IDENTITY`

---

## 4. Licensed next

Do not promote \(6\pi^2\). Do not enlarge the factor dictionary to chase \(0.20\%\). FB-O1′ remains **CLOSED F**. Pair EOM parked. \(T(2,3)\) not proton mass. \(\alpha\) INPUT. \(a\) INPUT. Do not name \(R\). No hub dump.

---

## 5. One-line for agents

> **§LSTAR-RX: reverse-engineer exception found no non-circular identity for \(12.48\,\mathrm{fm}\); \(6\pi^2\) is a \(0.20\%\) near-miss, not a proof; chips OPEN; no promote.**