Position: Contextual Integrity is Inadequately Applied to Language Models

Authors: Yan Shvartzshnaider, Vasisht Duddu

ICML 2025 | Venue PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Theoretical This position paper argues that existing literature inadequately applies CI for LLMs without embracing the theory s fundamental tenets.
Researcher Affiliation Academia 1York University 2University of Waterloo. Correspondence to: Yan Shvartzshnaider <EMAIL>, Vasisht Duddu <EMAIL>.
Pseudocode No The paper describes theoretical concepts and critiques methodologies, but does not present any structured pseudocode or algorithm blocks.
Open Source Code No The paper is a position paper and does not describe a methodology for which open-source code would typically be provided or referenced.
Open Datasets No The paper is a position paper that critiques existing research on Contextual Integrity for LLMs and does not present new experimental results requiring a specific dataset.
Dataset Splits No As the paper does not present new experimental results, there is no mention of dataset splits for training, validation, or testing.
Hardware Specification No The paper focuses on theoretical arguments and critiques of existing literature; it does not describe experimental procedures that would require specific hardware specifications.
Software Dependencies No The paper is a theoretical position paper and does not detail any experimental implementation requiring specific software dependencies or their version numbers.
Experiment Setup No The paper presents a theoretical position and critique of existing work, rather than conducting new experiments, thus no experimental setup details such as hyperparameters are provided.