Run · Voynich Manuscript

T42 generation 3: truncated-prior NPE round 2 (300 new draws per section); the round-1 gain over T39 does not hold

Follows [@deistler2022truncated]. A first launch failed on a scoping bug before any simulation; the script was fixed and rerun unchanged otherwise.

Calibration holds, but more simulations move the posterior mode and its fit falls back to T39's level: the round-1 improvement was within the posterior's spread, not a robust better setting. What is robust across both rounds and both sections: stream spacing, no syllable signs, no positional sign forms, the line layer present, simplified spelling. Nomenclator size, homophone counts and copied nulls are not pinned down by these 19 statistics. The remaining misses are the same in every fit: glyph predictability (h2), one-off words (hapax) and form-function coupling in Herbal A; line-position measures in Herbal B.coverage at least 0.954 for every setting in both sections. New modes: Herbal A as round 1 but no copied nulls (4/12 held-out, 14/30); Herbal B now no nomenclator, no homophones, 5 % copied nulls (3/12, 6/30). The 99.9 % region kept 46 of 47 setting values for A and 42 for B0

Result

Run	Measured	Result	Against	Decision	Date
Run	Out-of-fold 90 % coverage in the truncated region; held-out statistics missed by the new posterior mode (10 seeds, T39 scoring)	coverage at least 0.954 for every setting in both sections. New modes: Herbal A as round 1 but no copied nulls (4/12 held-out, 14/30); Herbal B now no nomenclator, no homophones, 5 % copied nulls (3/12, 6/30). The 99.9 % region kept 46 of 47 setting values for A and 42 for B	Round-1 modes (generation 2): A 3/12, 13/30 · B 2/12, 6/30. T39: A 5/12, 15/30 · B 3/12, 9/30	Calibration holds, but more simulations move the posterior mode and its fit falls back to T39's level: the round-1 improvement was within the posterior's spread, not a robust better setting. What is robust across both rounds and both sections: stream spacing, no syllable signs, no positional sign forms, the line layer present, simplified spelling. Nomenclator size, homophone counts and copied nulls are not pinned down by these 19 statistics. The remaining misses are the same in every fit: glyph predictability (h2), one-off words (hapax) and form-function coupling in Herbal A; line-position measures in Herbal B.

What it used and found

Neural posterior estimation for the period-cipher family, with simulation-based calibration

Method

MTH

Herbal A and Herbal B paragraph text: training and held-out pages

Dataset

DAT

Latin genre texts

Dataset

DAT

ZL3b transcription

Dataset

DAT

Latin as a re-spaced homophonic stream (period practices)

Answer it tests

ANS

Parameters

region = per setting, values inside the round-1 posterior's 99.9 % highest-posterior set, network = as round 1, new_draws = 300, sim_seeds = random.Random(seed*100 + section), confirm_seeds = 5000-5009

Cite it

Mysteries Department. (2026). T42 generation 3: truncated-prior NPE round 2 (300 new draws per section); the round-1 gain over T39 does not hold [Run RUN0DCG56DG7RVQ]. https://mysteriesdept.com/r/RUN0DCG56DG7RVQ

@misc{mysteriesdept_run0dcg56dg7rvq,
  title     = {T42 generation 3: truncated-prior NPE round 2 (300 new draws per section); the round-1 gain over T39 does not hold},
  author    = {{Mysteries Department}},
  year      = {2026},
  note      = {Run, catalog key RUN0DCG56DG7RVQ},
  url       = {https://mysteriesdept.com/r/RUN0DCG56DG7RVQ}
}