<?xml version="1.0" encoding="UTF-8"?><urlset xmlns="http://www.sitemaps.org/schemas/sitemap/0.9"><url><loc>https://mbrenndoerfer.com/writing/electra-efficient-pretraining-replaced-token-detection/electra</loc><lastmod>2025-07-25T00:00:00.000Z</lastmod></url><url><loc>https://mbrenndoerfer.com/writing/electra-efficient-pretraining-replaced-token-detection/the-inefficiency-of-masked-language-modeling</loc><lastmod>2025-07-25T00:00:00.000Z</lastmod></url><url><loc>https://mbrenndoerfer.com/writing/electra-efficient-pretraining-replaced-token-detection/the-generator-discriminator-architecture</loc><lastmod>2025-07-25T00:00:00.000Z</lastmod></url><url><loc>https://mbrenndoerfer.com/writing/electra-efficient-pretraining-replaced-token-detection/generator-training</loc><lastmod>2025-07-25T00:00:00.000Z</lastmod></url><url><loc>https://mbrenndoerfer.com/writing/electra-efficient-pretraining-replaced-token-detection/discriminator-training</loc><lastmod>2025-07-25T00:00:00.000Z</lastmod></url><url><loc>https://mbrenndoerfer.com/writing/electra-efficient-pretraining-replaced-token-detection/the-replaced-token-detection-objective</loc><lastmod>2025-07-25T00:00:00.000Z</lastmod></url><url><loc>https://mbrenndoerfer.com/writing/electra-efficient-pretraining-replaced-token-detection/joint-training</loc><lastmod>2025-07-25T00:00:00.000Z</lastmod></url><url><loc>https://mbrenndoerfer.com/writing/electra-efficient-pretraining-replaced-token-detection/why-no-adversarial-training</loc><lastmod>2025-07-25T00:00:00.000Z</lastmod></url><url><loc>https://mbrenndoerfer.com/writing/electra-efficient-pretraining-replaced-token-detection/sample-efficiency</loc><lastmod>2025-07-25T00:00:00.000Z</lastmod></url><url><loc>https://mbrenndoerfer.com/writing/electra-efficient-pretraining-replaced-token-detection/electra-scaling</loc><lastmod>2025-07-25T00:00:00.000Z</lastmod></url><url><loc>https://mbrenndoerfer.com/writing/electra-efficient-pretraining-replaced-token-detection/weight-sharing</loc><lastmod>2025-07-25T00:00:00.000Z</lastmod></url><url><loc>https://mbrenndoerfer.com/writing/electra-efficient-pretraining-replaced-token-detection/fine-tuning-electra</loc><lastmod>2025-07-25T00:00:00.000Z</lastmod></url><url><loc>https://mbrenndoerfer.com/writing/electra-efficient-pretraining-replaced-token-detection/using-pre-trained-electra</loc><lastmod>2025-07-25T00:00:00.000Z</lastmod></url><url><loc>https://mbrenndoerfer.com/writing/electra-efficient-pretraining-replaced-token-detection/performance-comparison</loc><lastmod>2025-07-25T00:00:00.000Z</lastmod></url><url><loc>https://mbrenndoerfer.com/writing/electra-efficient-pretraining-replaced-token-detection/limitations-and-impact</loc><lastmod>2025-07-25T00:00:00.000Z</lastmod></url><url><loc>https://mbrenndoerfer.com/writing/electra-efficient-pretraining-replaced-token-detection/key-parameters</loc><lastmod>2025-07-25T00:00:00.000Z</lastmod></url><url><loc>https://mbrenndoerfer.com/writing/electra-efficient-pretraining-replaced-token-detection/summary</loc><lastmod>2025-07-25T00:00:00.000Z</lastmod></url><url><loc>https://mbrenndoerfer.com/writing/electra-efficient-pretraining-replaced-token-detection/quiz</loc><lastmod>2025-07-25T00:00:00.000Z</lastmod></url></urlset>