LLMON: An LLM-native Markup Language to Leverage Structure and Semantics at the LLM Interface

TL;DR

This paper introduces LLMON, a markup language designed to encode structure and semantics in prompts, improving LLM understanding, safety, and security during training and inference.

Contribution

The paper presents the design and implementation of LLMON, an LLM-native markup language that enhances communication of structure and semantics to LLMs.

Findings

Preliminary evidence shows LLMON improves model accuracy.

LLMON enhances safety and security in LLM applications.

The approach enables better training and inference strategies.

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.