--- title: Writing seen in PHerc. 306B, for the first time in 2000... description: Liam Warner and Nicola Bodill won the $50,000 First Letters Prize for reading text in the Herculaneum scroll PHerc. 306B. The submission supports the reading... canonical: https://daily.dev/posts/writing-seen-in-pherc-306b-for-the-first-time-in-2000-years-ojj8dq5cq twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: Writing seen in PHerc. 306B, for the first time in 2000 years | daily.dev og:description: Liam Warner and Nicola Bodill won the $50,000 First Letters Prize for reading text in the Herculaneum scroll PHerc. 306B. The submission supports the reading... og:url: https://daily.dev/posts/writing-seen-in-pherc-306b-for-the-first-time-in-2000-years-ojj8dq5cq og:image: https://api.daily.dev/og/posts/ojJ8dq5cQ.png og:image:alt: Writing seen in PHerc. 306B, for the first time in 2000 years og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # Writing seen in PHerc. 306B, for the first time in 2000 years **[Hacker News](https://daily.dev/sources/hn)** · 8 min read · 2 upvotes · 0 comments ## Summary Liam Warner and Nicola Bodill won the $50,000 First Letters Prize for reading text in the Herculaneum scroll PHerc. 306B. The submission supports the reading 'τὸ δι᾽ αὑτὸ αἱρετόν' ('that which is in itself desirable'), a phrase with parallels in Philodemus's treatise On Anger. The prize required 10 readable characters in a 4 cm² area; 25 submissions were reviewed before this one succeeded. Their pipeline combined two ink-detection models, TRACE v3B and a custom fine-tuned model called A-hn, built on code and data they open-sourced. 20 more scanned scrolls remain eligible for First Letters Prizes, and this result makes PHerc. 306B a front-runner for the $800,000 Grand Prize. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### What Greek phrase was recovered from the Herculaneum scroll PHerc. 306B? The phrase read is 'τὸ δι᾽ αὑτὸ αἱρετόν', meaning 'that which is in itself desirable.' It parallels expressions found in Aristotle, Sextus Empiricus, and Philodemus's treatise On Anger, where he discusses things pursued as worthy in themselves. In an Epicurean context, this expression typically refers only to good and pleasure. _Developers curious how machine learning decodes carbonized scrolls can follow such breakthroughs on daily.dev._ ### How did Liam Warner and Nicola Bodill detect ink in the Vesuvius Challenge scroll PHerc. 306B? They multiplied the outputs of two ink-detection models so a pixel registered as ink only when both agreed. One model was TRACE v3B, a public model by Jashanjot Singh Gill and Nhat Nam Tran; the other was A-hn, a custom model they fine-tuned from Youssef Nader's v8-in model, trained with an extra loss penalty for marking blank papyrus as ink. _Anyone weighing model-ensembling strategies for detection tasks can track approaches like this on daily.dev._ ### What is the prize money for the Vesuvius Challenge Grand Prize for reading an entire scroll? The Grand Prize is $800,000, awarded to the first team that virtually unwraps and reads 70% of an entire Herculaneum scroll with minimal human input. The $50,000 First Letters Prize, by contrast, requires only 10 readable characters within a 4 cm² area, and 20 more scanned scrolls remain eligible for that smaller prize. _Teams tracking open science competitions and reward milestones can follow updates like this via daily.dev._ ## Community take How the wider developer community reacted, aggregated from 1 discussion and 12 comments across hackernews (as of 2026-11-03). **TL;DR:** Commenters are awestruck that legible text could be recovered from 2000-year-old carbonized scrolls, with reflections on past excavation losses and optimism about applying similar tech to remaining scrolls. **Sentiment:** 75% positive · 20% mixed · 5% skeptical **The case for** - Several note it's remarkable that any text could be reconstructed from carbonized, physically damaged scrolls. - Some draw an optimistic parallel to the human genome project, expecting a rapid acceleration once techniques mature. - A few appreciate the historical foresight of preserving the carbonized logs centuries ago on the hope they'd one day be readable. **The pushback** - Some lament that an unknown number of scrolls were destroyed or burned as firewood during early excavations, mistaken for ordinary logs. - A couple of lighthearted comments speculate the recovered content might 'just' be more Epicurean philosophy, downplaying its significance. **By community** - hackernews (positive): Enthusiastic reactions to the achievement, mixed with historical reflection on scroll losses and jokes about the find's significance. **Open questions** - How many more of the thousands of unscanned/unopened scrolls could eventually be read with similar techniques? **Highlights** > I'm reminded of the human genome project here - in that case there was a geometric ramp-up that meant we went from 1 or 2% sequenced to 100% really quickly. This sort of doubling when you can get it is pretty magical. I'd like to imagine that we'll get something similar here sometime in the next ten years -- I understand there are (tens of?) thousands of these scrolls still to be scanned much less unrolled. Just an incredibly cool thing to be doing. > — [vessenes on hackernews](https://news.ycombinator.com/item?id=50356531) > An unknown number of them were also allegedly burned for heat as they were mistaken for wood logs during early excavations > — [runeb on hackernews · 1 comments](https://news.ycombinator.com/item?id=50354881) > Amazing work. Makes you wonder what else is out there holding a record of the past, with us too primitive to read it. > — [jl6 on hackernews · 1 comments](https://news.ycombinator.com/item?id=50355171) > It’s amazing that they can reconstruct any text at all from those scrolls. > — [efitz on hackernews](https://news.ycombinator.com/item?id=50353654) **Source threads** - [hackernews](https://news.ycombinator.com/item?id=50342119) · 79 points · 12 comments ## Similar posts on daily.dev - [Some recent discoveries](https://daily.dev/posts/some-recent-discoveries-wnlc1k01n) · Arthur O'Dwyer · 0 upvotes · 0 comments - [Scientists develop new method for deciphering ancient scrolls](https://daily.dev/posts/scientists-develop-new-method-for-deciphering-ancient-scrolls-9ajev4aad) · Ars Technica · 0 upvotes · 0 comments - [The Writing Is on the Wall for Handwriting Recognition](https://daily.dev/posts/the-writing-is-on-the-wall-for-handwriting-recognition-30jmcibco) · Hacker News · 1 upvotes · 0 comments - [France’s Goncourt prize drops a novel over claims AI wrote it](https://daily.dev/posts/france-s-goncourt-prize-drops-a-novel-over-claims-ai-wrote-it-evfdwpwry) · The Next Web · 0 upvotes · 0 comments --- Tags: [#machine-learning](https://daily.dev/tags/machine-learning), [#open-source](https://daily.dev/tags/open-source), [#computer-vision](https://daily.dev/tags/computer-vision) [View this post on daily.dev](https://daily.dev/posts/writing-seen-in-pherc-306b-for-the-first-time-in-2000-years-ojj8dq5cq) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"Writing seen in PHerc. 306B, for the first time in 2000 years","url":"https://daily.dev/posts/writing-seen-in-pherc-306b-for-the-first-time-in-2000-years-ojj8dq5cq","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/writing-seen-in-pherc-306b-for-the-first-time-in-2000-years-ojj8dq5cq"},"datePublished":"2026-11-03T16:17:14.317Z","dateModified":"2026-11-03T19:48:12.123Z","description":"Liam Warner and Nicola Bodill won the $50,000 First Letters Prize for reading text in the Herculaneum scroll PHerc. 306B. The submission supports the reading...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/f9b0ba95c1ea455cc01bf15b2c7a6b0b?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/f9b0ba95c1ea455cc01bf15b2c7a6b0b?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"Hacker News","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"Hacker News","logo":"https://media.daily.dev/image/upload/t_logo,f_auto/v1/logos/hn","url":"https://daily.dev/sources/hn"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/writing-seen-in-pherc-306b-for-the-first-time-in-2000-years-ojj8dq5cq","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":2},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"machine-learning,open-source,computer-vision","timeRequired":"PT8M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Hacker News","item":"https://daily.dev/sources/hn"},{"@type":"ListItem","position":3,"name":"Writing seen in PHerc. 306B, for the first time in 2000 years"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/writing-seen-in-pherc-306b-for-the-first-time-in-2000-years-ojj8dq5cq#faq","mainEntity":[{"@type":"Question","name":"What Greek phrase was recovered from the Herculaneum scroll PHerc. 306B?","acceptedAnswer":{"@type":"Answer","text":"The phrase read is 'τὸ δι᾽ αὑτὸ αἱρετόν', meaning 'that which is in itself desirable.' It parallels expressions found in Aristotle, Sextus Empiricus, and Philodemus's treatise On Anger, where he discusses things pursued as worthy in themselves. In an Epicurean context, this expression typically refers only to good and pleasure. Developers curious how machine learning decodes carbonized scrolls can follow such breakthroughs on daily.dev."}},{"@type":"Question","name":"How did Liam Warner and Nicola Bodill detect ink in the Vesuvius Challenge scroll PHerc. 306B?","acceptedAnswer":{"@type":"Answer","text":"They multiplied the outputs of two ink-detection models so a pixel registered as ink only when both agreed. One model was TRACE v3B, a public model by Jashanjot Singh Gill and Nhat Nam Tran; the other was A-hn, a custom model they fine-tuned from Youssef Nader's v8-in model, trained with an extra loss penalty for marking blank papyrus as ink. Anyone weighing model-ensembling strategies for detection tasks can track approaches like this on daily.dev."}},{"@type":"Question","name":"What is the prize money for the Vesuvius Challenge Grand Prize for reading an entire scroll?","acceptedAnswer":{"@type":"Answer","text":"The Grand Prize is $800,000, awarded to the first team that virtually unwraps and reads 70% of an entire Herculaneum scroll with minimal human input. The $50,000 First Letters Prize, by contrast, requires only 10 readable characters within a 4 cm² area, and 20 more scanned scrolls remain eligible for that smaller prize. Teams tracking open science competitions and reward milestones can follow updates like this via daily.dev."}}]} ```