What does Provenance and Fidelity of Derivatives conclude, in numbers, about “The Compounding of Drift: Serial Reproduction, p to the Power k, and Repair”?

Answer: Under independent errors that are never repaired, a relay with fidelity 0.95 per layer keeps only 0.7738 of its facts at the fifth layer; that figure is a warning, not a law. With any repair at all a long chain settles at a positive level, but repair itself requires going back to the book. Drift also tends to have a direction, which averaging cannot remove. The question left open is what exactly should be compared when going back to the book, and in what sense a piece of text should count as the same as the master. The next chapter answers with provenance hashes.

Scope of the numbers: Parameters are illustrative and costs are accounting points, not prices; all calculations are generated by the book's programs. It is not individual legal or business advice; its "copyright chain" means only provenance and permission records for the author's own works, and publishing any derivative needs sign-off by the person in charge.

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: 0.7738, 0.95

share of facts intact after k layers (independent errors)
k p0.99 p=0.97 p=0.95 p=0.90
10.9900 0.9700 0.9500 0.9000
20.9801 0.9409 0.9025 0.8100
30.9703 0.9127 0.8574 0.7290
40.9606 0.8853 0.8145 0.6561
50.9510 0.8587 0.7738 0.5905
60.9415 0.8330 0.7351 0.5314
expected intact facts of 40 at layer 5:
38.0434.35 30.95 23.62
chance all 40 facts survive layer 3:
0.29940.0259 0.0021 0.0000
layers until the intact share drops below 0.80:
238 5 3
Show the full program
# Chapter 1. p11: fidelity compounds as p**k along a chain of derivatives.
# Model: a master book carries N = 40 checkable facts. Each compilation
# step (book -> summary -> script -> slides -> quiz -> clip) copies each
# fact correctly with probability p, independently, and never repairs
# an earlier error. Layer k then keeps a fact intact with probability
# p**k, and keeps all N facts intact with probability p**(N*k).
# The p values are illustrative, not measured.
import math

N = 40
PS = (0.99, 0.97, 0.95, 0.90)
print("share of facts intact after k layers (independent errors)")
print("k  " + "".join("   p=%.2f" % p for p in PS))
for k in range(1, 7):
    print("%d  " % k + "".join("   %6.4f" % (p ** k) for p in PS))
print("expected intact facts of %d at layer 5:" % N)
print("   " + "".join("   %6.2f" % (N * p ** 5) for p in PS))
print("chance all %d facts survive layer 3:" % N)
print("   " + "".join("   %6.4f" % (p ** (N * 3)) for p in PS))
print("layers until the intact share drops below 0.80:")
cut = [math.floor(math.log(0.80) / math.log(p)) + 1 for p in PS]
print("   " + "".join("   %6d" % c for c in cut))

Single source and citation

Fan, Y. (2026). Multimodal Derivative Compilation and Provenance Chains: A Structure-Preserving Pipeline from One Monograph to Podcasts, Slides and Quiz Banks (Version 1.0.0), Chapter 1: The Compounding of Drift: Serial Reproduction, p to the Power k, and Repair. 5D Library. DOI pending

DOI pending · monograph page goes live after publication · 5DL-B-268 · v1.0.0 · CC-BY-NC-ND-4.0

BibTeX
@book{fan2026268,
  author    = {Fan, Yafu},
  title     = {Multimodal Derivative Compilation and Provenance Chains: A Structure-Preserving Pipeline from One Monograph to Podcasts, Slides and Quiz Banks},
  year      = {2026},
  publisher = {5D Library},
  edition   = {1.0.0},
  note      = {DOI pending}
}

Other chapter answers in this book