The rapid emergence of new entities -- driven by cultural shifts, evolving trends, and personalized user data -- poses a significant challenge for existing Speech Large Language Models (Speech LLMs). While these models excel at general conversational tasks, their static training knowledge limits their ability to recognize domain-specific terms such as contact names, playlists, or technical jargon. Existing solutions primarily rely on prompting, which suffers from poor scalability: as the entity list grows, prompting encounters context window limitations, increased inference latency, and the "lost-in-the-middle" phenomenon. An alternative approach, Generative Error Correction (GEC), attempts to rewrite transcripts via post-processing but frequently suffers from "over-correction", introducing hallucinations of entities that were never spoken. In this work, we introduce LOGIC (Logit-Space Integration for Contextual Biasing), an efficient and robust framework that operates directly in the decoding layer. Unlike prompting, LOGIC decouples context injection from input processing, ensuring constant-time complexity relative to prompt length. Extensive experiments using the Phi-4-MM model across 11 multilingual locales demonstrate that LOGIC achieves an average 9% relative reduction in Entity WER with a negligible 0.30% increase in False Alarm Rate.

翻译：新实体的快速涌现——由文化变迁、趋势演变和个性化用户数据驱动——对现有语音大语言模型构成了重大挑战。尽管这些模型在通用对话任务上表现出色，但其静态训练知识限制了识别领域特定术语（如联系人姓名、播放列表或技术术语）的能力。现有解决方案主要依赖提示方法，但该方法存在可扩展性不足的问题：随着实体列表增长，提示会遭遇上下文窗口限制、推理延迟增加以及“中间丢失”现象。另一种生成式错误修正方法试图通过后处理重写转录文本，但常出现“过度修正”问题，引入从未被说出的实体幻觉。本研究提出LOGIC（面向上下文偏置的Logit空间集成），这是一个直接在解码层运行的高效鲁棒框架。与提示方法不同，LOGIC将上下文注入与输入处理解耦，确保相对于提示长度的恒定时间复杂度。基于Phi-4-MM模型在11个多语言区域进行的广泛实验表明，LOGIC实现了实体词错误率平均相对降低9%，同时误报率仅增加0.30%。