<?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>AI Memory on Antonio V. Franco</title><link>https://antoniovfranco.com/tags/ai-memory/</link><description>Recent content in AI Memory on Antonio V. Franco</description><generator>Hugo</generator><language>en</language><lastBuildDate>Sun, 03 May 2026 00:00:00 -0300</lastBuildDate><atom:link href="https://antoniovfranco.com/tags/ai-memory/index.xml" rel="self" type="application/rss+xml"/><item><title>Can Memory Make an AI Worse? My Benchmark with Qwen3.5-9B and Stellar Classification</title><link>https://antoniovfranco.com/posts/2026/memory-benchmark-qwen-stellar/</link><pubDate>Sun, 03 May 2026 00:00:00 -0300</pubDate><guid>https://antoniovfranco.com/posts/2026/memory-benchmark-qwen-stellar/</guid><description>Benchmark reveals naive ReasoningBank memory reduces accuracy by 22pp on SDSS stellar classification. RB-EC with consolidation, pruning, and class-balanced retrieval restores baseline performance.</description></item></channel></rss>