{"type": "article", "title": "LLMs Don't Have to Generate One Token at a Time: How Medusa and Multi-Token Prediction Cheat Autoregression", "publisher": "Web Pulse", "url": "https://wpnews.pro/news/llms-don-t-have-to-generate-one-token-at-a-time-how-medusa-and-multi-token-cheat", "original_source": "https://dev.to/shrsv/llms-dont-have-to-generate-one-token-at-a-time-how-medusa-and-multi-token-prediction-cheat-8ej", "published": "2026-09-03T18:28:00+00:00", "accessed": "2026-09-03", "id": "llms-don-t-have-to-generate-one-token-at-a-time-how-medusa-and-multi-token-cheat"}