<?xml version="1.0" encoding="UTF-8"?><urlset xmlns="http://www.sitemaps.org/schemas/sitemap/0.9"><url><loc>https://mbrenndoerfer.com/writing/multi-query-attention-memory-efficient-inference/multi-query-attention</loc><lastmod>2025-08-04T00:00:00.000Z</lastmod></url><url><loc>https://mbrenndoerfer.com/writing/multi-query-attention-memory-efficient-inference/the-kv-cache-memory-problem</loc><lastmod>2025-08-04T00:00:00.000Z</lastmod></url><url><loc>https://mbrenndoerfer.com/writing/multi-query-attention-memory-efficient-inference/multi-query-attention-extreme-sharing</loc><lastmod>2025-08-04T00:00:00.000Z</lastmod></url><url><loc>https://mbrenndoerfer.com/writing/multi-query-attention-memory-efficient-inference/implementing-multi-query-attention</loc><lastmod>2025-08-04T00:00:00.000Z</lastmod></url><url><loc>https://mbrenndoerfer.com/writing/multi-query-attention-memory-efficient-inference/mqa-memory-benefits</loc><lastmod>2025-08-04T00:00:00.000Z</lastmod></url><url><loc>https://mbrenndoerfer.com/writing/multi-query-attention-memory-efficient-inference/quality-trade-offs</loc><lastmod>2025-08-04T00:00:00.000Z</lastmod></url><url><loc>https://mbrenndoerfer.com/writing/multi-query-attention-memory-efficient-inference/mqa-vs-gqa</loc><lastmod>2025-08-04T00:00:00.000Z</lastmod></url><url><loc>https://mbrenndoerfer.com/writing/multi-query-attention-memory-efficient-inference/practical-implementation-considerations</loc><lastmod>2025-08-04T00:00:00.000Z</lastmod></url><url><loc>https://mbrenndoerfer.com/writing/multi-query-attention-memory-efficient-inference/summary</loc><lastmod>2025-08-04T00:00:00.000Z</lastmod></url><url><loc>https://mbrenndoerfer.com/writing/multi-query-attention-memory-efficient-inference/key-parameters</loc><lastmod>2025-08-04T00:00:00.000Z</lastmod></url><url><loc>https://mbrenndoerfer.com/writing/multi-query-attention-memory-efficient-inference/quiz</loc><lastmod>2025-08-04T00:00:00.000Z</lastmod></url></urlset>