<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/">
  <channel>
    <title>World-Models on Quang Minh Dinh</title>
    <link>https://quangminhdinh.github.io/tags/world-models/</link>
    <description>Recent content in World-Models on Quang Minh Dinh</description>
    <generator>Hugo</generator>
    <language>en-us</language>
    <lastBuildDate>Mon, 22 Jun 2026 00:00:00 +0000</lastBuildDate>
    <atom:link href="https://quangminhdinh.github.io/tags/world-models/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>A shared channel is not a shared world</title>
      <link>https://quangminhdinh.github.io/garden/shared-channel-not-shared-world/</link>
      <pubDate>Mon, 22 Jun 2026 00:00:00 +0000</pubDate>
      <guid>https://quangminhdinh.github.io/garden/shared-channel-not-shared-world/</guid>
      <description>Generative agents coordinate through language and memory, but they never share a world. That ceiling is the whole point.</description>
    </item>
    <item>
      <title>A world in a grain of sand</title>
      <link>https://quangminhdinh.github.io/garden/world-in-a-grain-of-sand/</link>
      <pubDate>Mon, 22 Jun 2026 00:00:00 +0000</pubDate>
      <guid>https://quangminhdinh.github.io/garden/world-in-a-grain-of-sand/</guid>
      <description>Blake&amp;#39;s image, read through world models: the whole inferred from the detail, both as a design principle and as evidence the model worked.</description>
    </item>
    <item>
      <title>Royal Road, before I had the words</title>
      <link>https://quangminhdinh.github.io/garden/why-royal-road-fascinates-me/</link>
      <pubDate>Mon, 22 Jun 2026 00:00:00 +0000</pubDate>
      <guid>https://quangminhdinh.github.io/garden/why-royal-road-fascinates-me/</guid>
      <description>Legendary Moonlight Sculptor is one of my favorite novels, and the things that hooked me as a kid turned out to be my research.</description>
    </item>
    <item>
      <title>The best stories in Royal Road were never written</title>
      <link>https://quangminhdinh.github.io/garden/stories-nobody-wrote/</link>
      <pubDate>Mon, 22 Jun 2026 00:00:00 +0000</pubDate>
      <guid>https://quangminhdinh.github.io/garden/stories-nobody-wrote/</guid>
      <description>What makes the world feel infinite isn&amp;#39;t more content, it&amp;#39;s emergence: outcomes no designer anticipated, born from player-world synergy.</description>
    </item>
    <item>
      <title>What would it actually take to build Royal Road?</title>
      <link>https://quangminhdinh.github.io/garden/building-royal-road/</link>
      <pubDate>Mon, 22 Jun 2026 00:00:00 +0000</pubDate>
      <guid>https://quangminhdinh.github.io/garden/building-royal-road/</guid>
      <description>Through a researcher&amp;#39;s lens: the properties that make Royal Road great, how close today&amp;#39;s LLM-driven games get, and the gap that remains.</description>
    </item>
    <item>
      <title>When discovering the action space is the game</title>
      <link>https://quangminhdinh.github.io/garden/action-space-as-discovery/</link>
      <pubDate>Mon, 22 Jun 2026 00:00:00 +0000</pubDate>
      <guid>https://quangminhdinh.github.io/garden/action-space-as-discovery/</guid>
      <description>Text adventures let you guess at an open verb set; modern games hand you a closed one. The same tension shows up in agent action spaces.</description>
    </item>
    <item>
      <title>Shared state is a tiny world model</title>
      <link>https://quangminhdinh.github.io/garden/shared-state-is-a-tiny-world-model/</link>
      <pubDate>Sun, 21 Jun 2026 00:00:00 +0000</pubDate>
      <guid>https://quangminhdinh.github.io/garden/shared-state-is-a-tiny-world-model/</guid>
      <description>A multiplayer game restated my research agenda in miniature: two agents can only coordinate when they agree on one authoritative model of the world.</description>
    </item>
    <item>
      <title>World models as shared substrate</title>
      <link>https://quangminhdinh.github.io/garden/world-models/</link>
      <pubDate>Tue, 10 Mar 2026 00:00:00 +0000</pubDate>
      <guid>https://quangminhdinh.github.io/garden/world-models/</guid>
      <description>Why a common representation of the world is upstream of believable multi-agent behaviour.</description>
    </item>
    <item>
      <title>Grounding signals beyond language</title>
      <link>https://quangminhdinh.github.io/garden/multi-agent-grounding/</link>
      <pubDate>Fri, 20 Feb 2026 00:00:00 +0000</pubDate>
      <guid>https://quangminhdinh.github.io/garden/multi-agent-grounding/</guid>
      <description>If we want agents to share more than what can be said, the grounding signal has to be richer than text.</description>
    </item>
    <item>
      <title>What even is a world state?</title>
      <link>https://quangminhdinh.github.io/garden/what-is-a-world-state/</link>
      <pubDate>Tue, 15 Aug 2023 00:00:00 +0000</pubDate>
      <guid>https://quangminhdinh.github.io/garden/what-is-a-world-state/</guid>
      <description>We talk about world models as if &amp;#39;the state&amp;#39; were given. The representation is the hard part.</description>
    </item>
  </channel>
</rss>
