关于「作品与辩护:有成本信号、累赘原理与廉价话语」,《零解释税与作品引力》给出的定量结论是什么?
结论:第五章比较作品与辩护两类信号:看 4 件可核验样本的判断准确率是 0.770,用自述达到同样准确率需要 66 句。
数据口径:书中参数一律为示意,全部计算由书中程序生成;本书是非个人化经营建议,不是心理咨询或临床判断,不作任何收益承诺,「显化」只指把构想做成可交付、可核验的作品。 第三章求出解释的盈亏阈值,示意参数下为 0.464,只对阈值以上受众解释的净值是 320.0,对所有人解释是负 203.8。
对照数据表
下表数字由本章 Python 程序原样打印,可重新运行复现。 程序只用 Python 标准库与结论吻合的打印值: 0.770, 66
| A. accuracy / matched found / bits, by number of items k | ||
| k | artifacts | talk |
| 1 | 0.700 / 0.000 / 0.056 | 0.700 / 0.000 / 0.003 |
| 2 | 0.735 / 0.490 / 0.107 | 0.700 / 0.000 / 0.007 |
| 4 | 0.770 / 0.652 / 0.199 | 0.700 / 0.000 / 0.014 |
| 8 | 0.831 / 0.552 / 0.348 | 0.700 / 0.000 / 0.027 |
| 16 | 0.907 / 0.825 / 0.549 | 0.704 / 0.185 / 0.053 |
| 32 | 0.965 / 0.931 / 0.746 | 0.725 / 0.367 / 0.102 |
| B. guessing 'unmatched' for all gives accuracy 0.700 | ||
| C. talk items needed to match 4 artifacts: 66 |
查看完整验算程序
# Chapter 5. p53: how well a reader sorts creators from k items.
# Prior share of creators matched to this reader: 0.3. A verifiable
# sample (artifact) looks matched with probability 0.70 if the creator
# is matched and 0.40 if not; a self-description (talk) "sounds matched"
# with probability 0.90 vs 0.85, since anyone can say it. The reader
# applies the Bayes rule to the count of positive items; we compute the
# exact accuracy and the share of matched creators found. Illustrative.
import math
PRIOR = 0.3
def pmf(k, x, p):
return math.comb(k, x) * p ** x * (1 - p) ** (k - x)
def sort_quality(k, p1, p0):
acc, found = [], []
for x in range(k + 1):
a = PRIOR * pmf(k, x, p1)
b = (1 - PRIOR) * pmf(k, x, p0)
if a > b:
acc.append(a)
found.append(a)
else:
acc.append(b)
return math.fsum(acc), math.fsum(found) / PRIOR
def bits(k, p1, p0):
terms = []
for x in range(k + 1):
a = PRIOR * pmf(k, x, p1)
b = (1 - PRIOR) * pmf(k, x, p0)
px = a + b
for joint, prior in ((a, PRIOR), (b, 1 - PRIOR)):
if joint > 0:
terms.append(joint * math.log2(joint / (px * prior)))
return math.fsum(terms)
if __name__ == "__main__":
print("A. accuracy / matched found / bits, by number of items k")
print(" k artifacts talk")
for k in (1, 2, 4, 8, 16, 32):
a1, f1 = sort_quality(k, 0.70, 0.40)
a2, f2 = sort_quality(k, 0.90, 0.85)
cell = "{:.3f} / {:.3f} / {:.3f}"
line = " {:>2} " + cell + " " + cell
print(line.format(k, a1, f1, bits(k, 0.70, 0.40),
a2, f2, bits(k, 0.90, 0.85)))
print("B. guessing 'unmatched' for all gives accuracy {:.3f}".format(
1 - PRIOR))
need = next(k for k in range(1, 400)
if sort_quality(k, 0.90, 0.85)[0] >=
sort_quality(4, 0.70, 0.40)[0])
print("C. talk items needed to match 4 artifacts: {}".format(need))唯一数据源与引用
范雅福(Fan, Y.). (2026). 《零解释税生存动力学:放弃低维认同求证、全功率自我显化与作品引力场母典》第 5 章「作品与辩护:有成本信号、累赘原理与廉价话语」(版本 1.0.0). 5D Library. DOI 待发布
DOI 待发布 · 专著页正式发布后上线(本站暂无专著页) · 5DL-B-432 · v1.0.0 · CC-BY-NC-ND-4.0
BibTeX
@book{fan2026432,
author = {Fan, Yafu},
title = {Zero Explanation-Tax Dynamics: Abandoning Validation-Seeking, Full-Power Reality Rendering and Proof-by-Artifact Gravity},
year = {2026},
publisher = {5D Library},
edition = {1.0.0},
note = {DOI pending}
}