让大模型生成可控的音频指令，确保零错误发送。
LLM4OSC: Profile-Bound Natural Language Control with Deterministic Validation for Open Sound Control
- 用设备配置文件约束语言模型输出，先生成结构化意图
- 通过确定性代码验证并编码，实现0%错误发送率
- 适合对可靠性要求高的现场演出和虚拟制作场景
开放声音控制（OSC）是专业音频、现场表演和虚拟制作中实时参数控制的主流通信协议。大语言模型虽能生成看似合理的OSC消息，但常出现地址幻觉、类型标签错误，且在语义改写下失效，这在演出关键场景中不可接受。本文提出LLM4OSC，一种本地优先架构：模型基于人工审核的设备配置文件，生成结构化意图JSON，再经确定性代码验证、钳位与编码后才发送UDP包。引入冻结评估框架，以错误发送率作为CI门控指标——即那些虽通过验证仍会错误发送的案例。在Max/MSP主配置（12种模式；8个字面表达+8个同义改写+4个拒绝案例）上，经配置标签增强、符号槽填充、自然语言优化及检索置信度门控后，所有后端B0–B3均通过冻结门控（100%语义准确率，0%错误发送）。其中B0（规则引擎）保持生产默认，延迟约0.05ms；LLM后端延迟约3–4秒。历史少样本方法B2在该套件上准确率从62.5%提升至100%，并非因0.5B模型本身变安全，而是得益于符号后处理。主张将‘提议-验证-发送’流程与错误发送率作为语言到控制系统的首要评估指标。
原文摘要 · Abstract (English)
Open Sound Control (OSC) is the dominant wire protocol for real-time parametric control in professional audio, live performance, and virtual production. Large language models can emit plausible OSC, but they hallucinate addresses, mishandle type tags, and fail under paraphrase- unacceptable in show-critical contexts. We present LLM4OSC, a local-first architecture in which models propose structured intent JSON over a human-reviewed device profile, and deterministic code validates, clamps, and encodes before any UDP send. We introduce a frozen evaluation harness with CI gates on wrong-send rate: mismatches that would still pass validation and transmit. On a Max/MSP hero profile (12 patterns; 8 literal + 8 paraphrase + 4 refusal cases), after profile tag enrichment, symbolic slot fill, NL refine, and a retrieval confidence gate, backends B0--B3 all pass frozen gates (100% semantic accuracy, 0% wrong-send). B0 (rules) remains the production default at ~0.05ms; LLM backends remain ~3-4s. Historical few-shot B2 accuracy of 62.5% rises to 100% on this suite only after symbolic post-processing- not because the 0.5B model alone becomes show-safe. We argue for propose-validate-send and wrong-send rate as first-class metrics for language-to-control systems.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。