What does Bicameral AI Architecture conclude, in numbers, about “The Intent Translation Protocol: Slots, Ambiguity Entropy and Upper Bounds”?

Answer: Chapter 4 counts ambiguity with a five-slot intent packet and its entropy, and rewrites "zero ambiguity" as an audit bound: 299 clean packets in a row are needed before the ambiguity rate can be stated below 0.010.

Scope of the numbers: Its parameters are illustrative; all calculations are generated by its programs; "bicameral" is an engineering metaphor and no model is claimed to have awareness; the book is not personalised business or engineering advice. Marked boxes separate published evidence, runnable Python calculations (23 programs), classical sources and the author's hypotheses.

中文版

Data table

The numbers below are printed verbatim by the chapter's Python program and can be reproduced by running it. standard-library Python onlyprinted values that match the answer: 0.010, 299

A. one-sided 95 % upper bound on the ambiguity rate
n audited x0 x=1 x=3 rule of three
300.0950 0.1486 0.2386 0.1000
1000.0295 0.0466 0.0757 0.0300
3000.0099 0.0157 0.0256 0.0100
1,0000.0030 0.0047 0.0077 0.0030
3,0000.0010 0.0016 0.0026 0.0010
B. clean audits needed to certify the rate below a target
target 0.05059 packets with x = 0
target 0.020149 packets with x = 0
target 0.010299 packets with x = 0
target 0.005598 packets with x = 0
target 0.0012,995 packets with x = 0
C. a claim of 'no ambiguity' after 50 clean packets means
only: rate below 0.0582 at 95 %
Show the full program
# Chapter 4. p43: "zero ambiguity" becomes an upper bound.
# An audit reads n translated packets and finds x that an executor
# could read two ways. The exact one-sided 95 % Clopper-Pearson upper
# bound solves P(Binomial(n, u) <= x) = 0.05 for u (bisection on the
# exact binomial CDF). For x = 0 the rule of three, 3 / n, is close.
# Also: the audit size needed so that x = 0 certifies u < target.
import math


def binom_cdf(x, n, p):
    terms = [math.comb(n, k) * p ** k * (1.0 - p) ** (n - k)
             for k in range(x + 1)]
    return math.fsum(terms)


def upper(x, n, alpha=0.05):
    lo, hi = x / n, 1.0
    for _ in range(80):
        mid = (lo + hi) / 2.0
        if binom_cdf(x, n, mid) > alpha:
            lo = mid
        else:
            hi = mid
    return hi


def n_for_zero(target, alpha=0.05):
    return math.ceil(math.log(alpha) / math.log(1.0 - target))


if __name__ == "__main__":
    print("A. one-sided 95 % upper bound on the ambiguity rate")
    print("   n audited    x=0      x=1      x=3      rule of three")
    for n in (30, 100, 300, 1000, 3000):
        print("   {:>6,}       {:.4f}   {:.4f}   {:.4f}   {:.4f}".format(
            n, upper(0, n), upper(1, n), upper(3, n), 3.0 / n))
    print("B. clean audits needed to certify the rate below a target")
    for t in (0.05, 0.02, 0.01, 0.005, 0.001):
        print("   target {:.3f}: {:>5,} packets with x = 0".format(
            t, n_for_zero(t)))
    print("C. a claim of 'no ambiguity' after 50 clean packets means")
    print("   only: rate below {:.4f} at 95 %".format(upper(0, 50)))

Single source and citation

Fan, Y. (2026). Bicameral AI Architecture: Fusing Long-Context Holographic Strategy Models with Local Deterministic Execution Agents (Version 1.0.0), Chapter 4: The Intent Translation Protocol: Slots, Ambiguity Entropy and Upper Bounds. 5D Library. DOI pending

DOI pending · monograph page goes live after publication · 5DL-B-410 · v1.0.0 · CC-BY-NC-ND-4.0

BibTeX
@book{fan2026410,
  author    = {Fan, Yafu},
  title     = {Bicameral AI Architecture: Fusing Long-Context Holographic Strategy Models with Local Deterministic Execution Agents},
  year      = {2026},
  publisher = {5D Library},
  edition   = {1.0.0},
  note      = {DOI pending}
}

Other chapter answers in this book