# Authority revision notes

These notes are for the next time the authority file is deliberately re-pinned.
`vendor/fsot_compute.py` is unchanged. Pin **AEB2AD**. Nothing below is a value edit.

`full_report()` prints 343 constants with targets and 342 within 5%. Read that 342 as two groups:

| Group | Within 5% | Outside 5% |
|-------|----------:|-----------:|
| Identity or definition | 8 | 0 |
| Seed-derived | 334 | 1 |
| Total | 342 | 1 |

The row outside 5% is `Gluon_condensate` (`C_cosm − e⁻³`), about 6.94%.

`DNA_base_pair` is one of the 334 seed-derived rows and uses a measured input. See B-02.

## B-01 · Identity or definition rows

The computed side is the target literal, or a known exact rational. These rows are inside the 342. They are not seed-derived predictions.

| Row | Line | What the file stores |
|-----|-----:|----------------------|
| `Proton_radius` | 418 | formula `"exact"`, `mpf("0.8413")` against `mpf("0.8413")` |
| `STDP_Tau_Plus_ms` | 924 | `mpf(20)` against `mpf(20)` |
| `STDP_Tau_Minus_ms` | 925 | `mpf(20)` against `mpf(20)` |
| `Metatron_Spheres` | 881 | `mpf(13)` against `mpf(13)` |
| `Metatron_Pathways` | 882 | `mpf(27)` against `mpf(27)`, formula `3³=27` |
| `Max_Trits` | 977 | `mpf(27)` against `mpf(27)`, formula `3³` |
| `Potts3_beta` | 754 | `mpf(1)/9` against `mpf("0.11111")`. 1/9 is the exact 2-D 3-state Potts exponent. The target is the five-digit literal. |
| `Quark_condensate` | 757 | `mpf(1)/4` against `mpf("0.250")` |

Other rows land on their target because a seed expression evaluates to that integer or short decimal (`floor(2π)` against 6, `(25/25)^0.2` against 1, `φ + 1/φ²` against 2). Those stay in the 334. The computed side is an expression, not a copy of the target literal.

## B-02 · Measured input

Line 398 sets `a0 = mpf("0.529177")`, commented as the Bohr radius in angstroms. Line 412 uses it:

`DNA_base_pair`: `v = 2*PI*a0*(1 + GAMMA/(PI**2 * E))` against `mpf("3.4")`.

Mark this row **uses measured input**. CODATA 2022 is 0.529177210544. The literal stops at six decimals, about 4×10⁻⁷ relative. The next revision should say so on the row.

## B-03 · Fifteen-digit cube-root exponents

The engine works at 50 decimal digits. These four exponents are the literal `mpf("0.333333333333333")`. Each differs from 1/3 by about 3.3×10⁻¹⁶. The formula text already says 1/3.

| Row | Line |
|-----|-----:|
| `Feigenbaum_delta` | 433 |
| `Feigenbaum_alpha` | 436 |
| `Apery_zeta3` | 439 |
| `Water_anomaly_C` | 442 |

If the next revision means a cube root, write `mpf(1)/3`.

The same question appears once more, with a shorter literal: `CMB_tau` at line 987 uses `mpf("0.333333333333")` (twelve 3s).

## B-05 · Version label

Line 4 of the docstring says `FSOT 2.0 — Complete Computational Engine`. The printed banner in `full_report()` (line 1194) says the same. The hub and the pin are FSOT 2.1. The next revision should say 2.1.

## G-02 · m_μ/m_e integers 35 and 145, and the twelve-digit exponent

No derivation of the integers 35 and 145 is recorded. `vendor/fsot_compute.py` line 814 is

`v3 = (35*PHI**(-5) + 145) * E**mpf("0.333333333333")`

and the formula string on the next line says `(35φ⁻⁵+145)·e^(1/3)`. The section title above `lepton_ratios` says "Independent Derivations" and does not derive either integer. A search of the docs and the other copies of this line (the two Kaggle bundles) finds the same expression and no derivation.

The exponent literal is twelve 3s. It is the same truncation as line 812 (`m_tau/m_mu_lepton`) and line 987 (`CMB_tau`). It differs from 1/3 by about 3.3×10⁻¹⁶. Because the base is e, replacing the literal with 1/3 moves the product by about 3.3×10⁻¹³ relative. That shift is below the Ledger A relative error against the anchor, about 1.4×10⁻⁶. The literal stays as written. This note does not change the formula.

`docs/DERIVATIONS.md` is not added. There is no derivation to write down.

## Left off this list

B-04 (older copies of `C_EFF` and `K` in other repos) and B-06 (the 50-digit `GAMMA` and `G_CAT` literals match mpmath) are context from the audit. They are not edits for this pin.