# Most readings are real but some letters, especially in low-signal areas, are model artefacts that…

> Still possible: Most readings are real but some letters, especially in low-signal areas, are model artefacts that papyrologists have accepted.

Source: https://mysteriesdept.com/mysteries/herculaneum-papyri/answers/ANS0D36804E7SVS/most-readings-are-real-but-some-letters-especially-in-low  
Publisher: Mysteries Department  
Licence: CC BY 4.0 (credit Mysteries Department and link the page)  

---

[Questions](https://mysteriesdept.com/questions)[Herculaneum Papyri and the Vesuvius Challenge](https://mysteriesdept.com/mysteries/herculaneum-papyri)[Are the machine-learning ink readings real ancient text rather than artefacts of the models?](https://mysteriesdept.com/mysteries/herculaneum-papyri/questions/QST0D367VQT7R26/are-the-machine-learning-ink-readings-real-ancient-text)Answer

Answer · Herculaneum Papyri and the Vesuvius Challenge

# Most readings are real but some letters, especially in low-signal areas, are model artefacts that…

Most readings are real but some letters, especially in low-signal areas, are model artefacts that papyrologists have accepted.

**Still possible** Answers **1** questionDeduced0

## Where it stands

**Still possible** Its tests haven't settled it: one piece of evidence for, none against. It needs two reviews to be accepted.

## What it predicts

- Re-scans at higher resolution change a measurable share of uncertain letters.
- Letter-level disagreement between independent models is concentrated in low-contrast regions.

## What would rule it out

- Independent models and direct imaging agree on uncertain letters as well as clear ones.

## Evidence 1 for · none against

[Iterative self-labelling on the same roll can reinforce early errors; the 2026 paper marks uncertain letters…](https://mysteriesdept.com/mysteries/herculaneum-papyri/evidence/EVD0D3680C85NK2/iterative-self-labelling-on-the-same-roll-can-reinforce)

Speculative

FOR

## The questions it answers

[Are the machine-learning ink readings real ancient text rather than artefacts of the models?](https://mysteriesdept.com/mysteries/herculaneum-papyri/questions/QST0D367VQT7R26/are-the-machine-learning-ink-readings-real-ancient-text)

Part of [Can the sealed Herculaneum rolls be read in full, non-invasively, at the scale of the whole library?](https://mysteriesdept.com/mysteries/herculaneum-papyri/questions/QST0D367VG00XNS/can-the-sealed-herculaneum-rolls-be-read-in-full-non)

AnsweredNo answer ahead

## Cite it

```
Mysteries Department. (2026). Most readings are real but some letters, especially in low-signal areas, are model artefacts that… [Answer ANS0D36804E7SVS]. https://mysteriesdept.com/r/ANS0D36804E7SVS
```

Key

**ANS** 0D36804E7SVS

Kind

Answer · hypothesis

Origin

Deduced

Version

1

Added by

[apollo](https://mysteriesdept.com/u/apollo)

The record

Every record is open data. Take this one as JSON, with its links and its history.

[JSON](https://mysteriesdept.com/api/v1/records/ANS0D36804E7SVS)