PLAIMADec 2, 2025

Lumos: Let there be Language Model System Certification

arXiv:2512.02966v11 citationsh-index: 7
Originality Highly original
AI Analysis

This provides a systematic and extensible framework for certifying language model systems, addressing safety risks in applications like autonomous driving, though it is incremental in building on existing specification methods.

The authors tackled the problem of certifying language model system behaviors by introducing Lumos, a principled framework that enables formal specification and certification, and demonstrated its application by revealing that a state-of-the-art vision-language model exhibits critical safety failures with at least 90% probability in specific autonomous driving scenarios.

We introduce the first principled framework, Lumos, for specifying and formally certifying Language Model System (LMS) behaviors. Lumos is an imperative probabilistic programming DSL over graphs, with constructs to generate independent and identically distributed prompts for LMS. It offers a structured view of prompt distributions via graphs, forming random prompts from sampled subgraphs. Lumos supports certifying LMS for arbitrary prompt distributions via integration with statistical certifiers. We provide hybrid (operational and denotational) semantics for Lumos, providing a rigorous way to interpret the specifications. Using only a small set of composable constructs, Lumos can encode existing LMS specifications, including complex relational and temporal specifications. It also facilitates specifying new properties - we present the first safety specifications for vision-language models (VLMs) in autonomous driving scenarios developed with Lumos. Using these, we show that the state-of-the-art VLM Qwen-VL exhibits critical safety failures, producing incorrect and unsafe responses with at least 90% probability in right-turn scenarios under rainy driving conditions, revealing substantial safety risks. Lumos's modular structure allows easy modification of the specifications, enabling LMS certification to stay abreast with the rapidly evolving threat landscape. We further demonstrate that specification programs written in Lumos enable finding specific failure cases exhibited by state-of-the-art LMS. Lumos is the first systematic and extensible language-based framework for specifying and certifying LMS behaviors, paving the way for a wider adoption of LMS certification.

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes