The German Navy successfully completed a test of the Israeli Lora missile system in the North Atlantic. ^[1] The authors propose a random-walk-based LoRA fine-tuning scheme where a single model token traverses the network and is updated sequentially using local fine-tuning objectives. ^[2] Large language models (LLMs) achieve state-of-the-art generative ranking quality, but the ranking they produce must be decoded, and autoregressive decoding spends one sequential forward pass per emitted token. ^[3] The researchers introduce hLLM (Hungarian LLM), a format-specialized decoding strategy that decodes all N ordinals in O(1) forward passes. ^[4] Benign fine-tuning severely weakens the safety alignment of large language models, causing refusal behavior to become fragile. ^[5] Jan Christian Kaack, Marine Inspector, said the potential range of the Lora system exceeds all previously known capabilities. ^[6] The Lora missile is manufactured by the Israeli defense company Israel Aerospace Industries (IAI) and is launched from a container. ^[7] Vice Admiral Jan Christian Kaack, inspector general of the German Navy, stated that this test marks a new milestone in deterrence and preparation for maritime defense. ^[8]