<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:dc="http://purl.org/dc/elements/1.1/">
  <channel>
    <title>OxARCA</title>
    <link>https://oxarca.github.io/index.html</link>
    <description>OxARCA builds Archos, an epistemically grounded LLM system for archival exploration: source-grounded synthesis with sentence-level traceability, preserved ambiguity and a cumulative research record.</description>
    <language>en</language>
    <atom:link href="https://oxarca.github.io/feed.xml" rel="self" type="application/rss+xml"/>
    <lastBuildDate>Mon, 24 Aug 2026 11:00:00 GMT</lastBuildDate>
    <item>
      <title>Archos at the Digital Humanities @ Oxford Summer School 2026</title>
      <link>https://oxarca.github.io/post/2026-08-dhoxss-poster.html</link>
      <guid isPermaLink="true">https://oxarca.github.io/post/2026-08-dhoxss-poster.html</guid>
      <pubDate>Mon, 24 Aug 2026 11:00:00 GMT</pubDate>
      <dc:creator>OxARCA</dc:creator>
      <category>News</category>
      <description>We are presenting the Archos poster at DHOxSS.</description>
      <content:encoded>&lt;p&gt;We are presenting Archos at the &lt;strong&gt;Digital Humanities @ Oxford Summer School&lt;/strong&gt; on 24 August 2026 at the University of Oxford.&lt;/p&gt;
&lt;p&gt;The poster — &lt;em&gt;Interdisciplinary Uses of AI in the Humanities: Developing an Epistemically
Grounded LLM-System for Archival Exploration&lt;/em&gt; — sets out the problem of the 'archival deluge'. &lt;/p&gt;
&lt;p&gt;In the test run presented in this poster, using the &lt;a href="https://newtonproject.ox.ac.uk/texts/newtons-works/mint"&gt;Newton's Royal Mint papers of 1692–1700&lt;/a&gt;, we took a single research question about William Chaloner's (1650 – 1699) prosecution. Archos identified 261 primary sources and produced 5 sections and 27 subsections, with
1,688 sentence-level citations across 85,789 words of source-grounded prose. Whole-run compute
came to roughly $12.70.&lt;/p&gt;
&lt;h2&gt;The poster&lt;/h2&gt;
&lt;p class="my-6"&gt;
&lt;a href="/poster_archos.pdf" class="inline-block rounded-md border border-oxford-300 px-5 py-2.5 text-sm font-medium text-oxford-800 no-underline hover:border-bronze-400 hover:text-bronze-500 focus-visible:outline-2 focus-visible:outline-offset-2 focus-visible:outline-bronze-500 dark:border-oxford-700 dark:text-parchment-200 dark:hover:border-bronze-300 dark:hover:text-bronze-300"&gt;Download the poster (PDF, 1.5&amp;nbsp;MB)&lt;/a&gt;
&lt;/p&gt;

&lt;p&gt;If you are at the summer school, come and find us. We are especially keen to talk to anyone with
a collection they would like to discover through the Archos system — see &lt;a href="/page/datasets.html"&gt;your collection&lt;/a&gt; for what that involves.&lt;/p&gt;</content:encoded>
    </item>
    <item>
      <title>British Academy/Leverhulme Small Research Grant awarded</title>
      <link>https://oxarca.github.io/post/2026-08-ba-leverhulme-srg.html</link>
      <guid isPermaLink="true">https://oxarca.github.io/post/2026-08-ba-leverhulme-srg.html</guid>
      <pubDate>Tue, 18 Aug 2026 11:00:00 GMT</pubDate>
      <dc:creator>OxARCA</dc:creator>
      <category>Awards</category>
      <description>The British Academy and the Leverhulme Trust have funded the project - Transferable and Cost-Aware Source-Grounded AI for Archival Research</description>
      <content:encoded>&lt;p&gt;Project &lt;em&gt;Transferable and Cost-Aware Source-Grounded AI for Archival Research&lt;/em&gt;  has been
awarded a &lt;strong&gt;British Academy/Leverhulme Small Research Grant&lt;/strong&gt; (SRG26\262244). The principal
investigators are David De Roure and Yishun Lu.&lt;/p&gt;
&lt;p&gt;Our thanks to the British Academy and the Leverhulme Trust. Details of the scheme are
&lt;a href="https://www.thebritishacademy.ac.uk/funding/ba-leverhulme-small-research-grants/"&gt;here&lt;/a&gt;.&lt;/p&gt;</content:encoded>
    </item>
    <item>
      <title>The project is funded by a CrossLinks Award 2026</title>
      <link>https://oxarca.github.io/post/2026-01-crosslinks-award.html</link>
      <guid isPermaLink="true">https://oxarca.github.io/post/2026-01-crosslinks-award.html</guid>
      <pubDate>Thu, 07 May 2026 11:00:00 GMT</pubDate>
      <dc:creator>OxARCA</dc:creator>
      <category>Awards</category>
      <description>St Cross College has funded the development of an epistemically grounded LLM system for archival exploration through its CrossLinks Award scheme.</description>
      <content:encoded>&lt;p&gt;Our project — &lt;em&gt;Interdisciplinary Uses of AI in the Humanities: Developing an Epistemically
Grounded LLM-System for Archival Exploration&lt;/em&gt; — has been funded by a &lt;strong&gt;CrossLinks Award 2026&lt;/strong&gt;
from St Cross College, University of Oxford.&lt;/p&gt;
&lt;p&gt;The question we are funded to answer is under which conditions AI systems can help historical scholarship. That work became &lt;a href="/page/archos.html"&gt;Archos&lt;/a&gt;.&lt;/p&gt;
&lt;p&gt;Our thanks to St Cross College. For more information about the CrossLinks Award scheme, see &lt;a href="https://www.stx.ox.ac.uk/crosslinks"&gt;here&lt;/a&gt; and the news &lt;a href="https://www.stx.ox.ac.uk/article/first-crosslinks-award-winners-announced"&gt;here&lt;/a&gt;.&lt;/p&gt;</content:encoded>
    </item>
  </channel>
</rss>
