<?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>Block Diffusion on Negi AI Lab</title><link>https://ai.negi-lab.com/tags/block-diffusion/</link><description>Recent content in Block Diffusion on Negi AI Lab</description><image><title>Negi AI Lab</title><url>https://ai.negi-lab.com/images/og-default.png</url><link>https://ai.negi-lab.com/images/og-default.png</link></image><generator>Hugo -- 0.154.5</generator><language>ja</language><lastBuildDate>Sat, 09 May 2026 14:53:58 +0900</lastBuildDate><atom:link href="https://ai.negi-lab.com/tags/block-diffusion/index.xml" rel="self" type="application/rss+xml"/><item><title>dflash 使い方と性能レビュー 推論速度を3倍にするBlock Diffusionの衝撃</title><link>https://ai.negi-lab.com/posts/dflash-block-diffusion-llm-inference-review/</link><pubDate>Sat, 09 May 2026 00:00:00 +0900</pubDate><guid>https://ai.negi-lab.com/posts/dflash-block-diffusion-llm-inference-review/</guid><description>推測デコードに拡散モデルの概念を導入し、LLMの自己回帰生成におけるボトルネックを根本から改善する。。従来のFlash Speculative Decod...</description></item></channel></rss>