arXiv cs.AI· Michael Hind, Basel Shbita, Bo Wu, Farhan Ahmed, Chad DeLuca, Nathan Fulton, David Cox, Dan Gutfreund·· 2 天前AI 评分22
LLMON：一种 LLM 原生标记语言，在 LLM 接口中利用结构与语义
LLMON: An LLM-native Markup Language to Leverage Structure and Semantics at the LLM Interface
AI 导读
研究者提出 LLM 原生标记语言 LLMON（LLM Object Notation），用于向 LLM 传递文本的结构与语义元数据，可应用于模型训练、提示和推理实现，以提升模型准确性、安全性与安全性。该工作讨论了 LLM 原生标记语言的设计要求，并给出 LLMON 在训练与推理两类用例上的初步实证证据。
正文 · 原文
Abstract:Textual Large Language Models (LLMs) provide a simple and familiar interface: a string of text is used for both input and output. However, the information conveyed to an LLM often has a richer structure and semantics, which is not conveyed in a string. For example, most prompts contain both instructions ("Summarize this paper into a paragraph") and data (the paper to summarize), but these are usually not distinguished when passed to the model. This can lead to model confusion and security risks, such as prompt injection attacks.
This work addresses this shortcoming by introducing an LLM-native mark-up language, LLMON (LLM Object Notation, pronounced "Lemon"), that enables the structure and semantic metadata of the text to be communicated in a natural way to an LLM. This information can then be used during model training, model prompting, and inference implementation, leading to improvements in model accuracy, safety, and security. This is analogous to how programming language types can be used for many purposes, such as static checking, code generation, dynamic checking, and IDE highlighting.
We discuss the general design requirements of an LLM-native markup language, introduce the LLMON markup language and show how it meets these design requirements, describe how the information contained in a LLMON artifact can benefit model training and inference implementation, and provide some preliminary empirical evidence of its value for both of these use cases. We also discuss broader issues and research opportunities that are enabled with an LLM-native approach.
| Comments: | 28 pages |
| Subjects: | Software Engineering (cs.SE); Artificial Intelligence (cs.AI); Programming Languages (cs.PL) |
| Cite as: | arXiv:2603.22519 [cs.SE] |
| (or arXiv:2603.22519v4 [cs.SE] for this version) | |
| https://doi.org/10.48550/arXiv.2603.22519 arXiv-issued DOI via DataCite |
Submission history
From: Michael Hind [view email]
[v1]
Mon, 23 Mar 2026 19:27:35 UTC (137 KB)
[v2]
Mon, 30 Mar 2026 20:39:09 UTC (135 KB)
[v3]
Tue, 13 Oct 2026 23:40:21 UTC (186 KB)
[v4]
Thu, 15 Oct 2026 02:24:57 UTC (186 KB)
来源：arXiv cs.AI · arxiv.org