What does Hard Gates and Soft-Prompt Decay conclude, in numbers, about “Classical Sources East and West: Law and the Inked Line in the Han Feizi, and Reason's Classification of Error”?

Answer: Chapter 7 examines the chapter "You Du" of the Han Feizi and Reason's classification of error; behind the gate, 0.627 of the residual is knowledge-based mistakes.

Scope of the numbers: Soft reminders work in proportion; when layers are independent the residual violation rate is the product of their miss rates, and with illustrative parameters four layers still leave 0.0048, not zero. Its parameters are illustrative; all calculations are generated by its programs.

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: 0.627

kind share c s after rev after gate
slip/lapse0.45 0.95 0.50 0.378 0.053
rule-based0.25 0.60 0.40 0.252 0.283
knowledge0.20 0.05 0.30 0.235 0.627
violation0.10 0.90 0.20 0.134 0.038
residual left: review only 0.5950, review plus gate 0.2122
latent hole: a2 a week, p = 0.50, escapes 4.0
latent hole: a2 a week, p = 0.10, escapes 20.0
latent hole: a10 a week, p = 0.10, escapes 100.0
Show the full program
# Chapter 7. p72: Reason's error types meet a pattern gate.
# Illustrative mix of errors in an agent's saved chapters, by the kind
# Reason (1990) distinguishes: skill-based slips and lapses, rule-based
# mistakes, knowledge-based mistakes, and violations of a standing
# rule. Each kind has a chance to leave a mark a code gate can match
# (c) and a chance to be caught by a soft review (s). The residual mix
# after review only and after review plus gate shows which kind is
# left. Then a latent hole in the rule set: escapes through it are
# a / p when a errors a week meet it and a weekly review finds the
# hole with probability p.
from fractions import Fraction as F

KINDS = [("slip/lapse", F(45, 100), F(95, 100), F(50, 100)),
         ("rule-based", F(25, 100), F(60, 100), F(40, 100)),
         ("knowledge", F(20, 100), F(5, 100), F(30, 100)),
         ("violation", F(10, 100), F(90, 100), F(20, 100))]
print("%-11s %6s %6s %6s %10s %10s" % ("kind", "share", "c", "s",
                                       "after rev", "after gate"))
r_tot, g_tot = F(0), F(0)
rows = []
for name, w, c, s in KINDS:
    rv = w * (1 - s)
    gt = rv * (1 - c)
    r_tot += rv
    g_tot += gt
    rows.append((name, w, c, s, rv, gt))
for name, w, c, s, rv, gt in rows:
    print("%-11s %6.2f %6.2f %6.2f %10.3f %10.3f" % (
        name, float(w), float(c), float(s), float(rv / r_tot),
        float(gt / g_tot)))
print("residual left: review only %.4f, review plus gate %.4f" % (
    float(r_tot), float(g_tot)))
for a, p in ((2, F(1, 2)), (2, F(1, 10)), (10, F(1, 10))):
    print("latent hole: a = %2d a week, p = %.2f, escapes %.1f" % (
        a, float(p), float(a / p)))

Single source and citation

Fan, Y. (2026). Hard Gates across the Client Lifecycle: the Decay of Soft-Prompt Compliance over Turns and State Machines for Interception before Publication (Version 1.0.0), Chapter 7: Classical Sources East and West: Law and the Inked Line in the Han Feizi, and Reason's Classification of Error. 5D Library. DOI pending

DOI pending · monograph page goes live after publication · 5DL-B-256 · v1.0.0 · CC-BY-NC-ND-4.0

BibTeX
@book{fan2026256,
  author    = {Fan, Yafu},
  title     = {Hard Gates across the Client Lifecycle: the Decay of Soft-Prompt Compliance over Turns and State Machines for Interception before Publication},
  year      = {2026},
  publisher = {5D Library},
  edition   = {1.0.0},
  note      = {DOI pending}
}

Other chapter answers in this book