<?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>Geography on Dylan Ler</title><link>http://dylanler.github.io/tags/geography/</link><description>Recent content in Geography on Dylan Ler</description><generator>Hugo -- 0.133.0</generator><language>en-us</language><lastBuildDate>Fri, 02 Oct 2026 01:20:00 -0700</lastBuildDate><atom:link href="http://dylanler.github.io/tags/geography/index.xml" rel="self" type="application/rss+xml"/><item><title>Blind Earth with Clef, Clef-flash, and Jev</title><link>http://dylanler.github.io/posts/blind-earth-clef-jev/</link><pubDate>Fri, 02 Oct 2026 01:20:00 -0700</pubDate><guid>http://dylanler.github.io/posts/blind-earth-clef-jev/</guid><description>Henry’s How Does A Blind Model See The Earth? asks a language model, cell by cell, whether a lat/lon is over land or water, then paints the answers on an equirectangular grid. No images go in. Whatever structure appears in the map is whatever geographic prior the model already carries.
This post adapts that recipe for System One decision models — Cloudflare Clef / Clef-flash and TypeSafe Jev — using a binary choice (Land vs Water) instead of free-form generation.</description></item></channel></rss>