Position: Uncertainty Quantification Needs Reassessment for Large Language Model Agents
Authors: Michael Kirchhof, Gjergji Kasneci, Enkelejda Kasneci
ICML 2025 | Venue PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Theoretical | This position paper argues that this traditional dichotomy of uncertainties is too limited for the open and interactive setup that LLM agents operate in when communicating with a user, and that we need to research avenues that enrich uncertainties in this novel scenario. We review the literature and find that popular definitions of aleatoric and epistemic uncertainties directly contradict each other and lose their meaning in interactive LLM agent settings. Hence, we propose three novel research directions that focus on uncertainties in such human-computer interactions: Underspecification uncertainties, for when users do not provide all information or define the exact task at the first go, interactive learning, to ask follow-up questions and reduce the uncertainty about the current context, and output uncertainties, to utilize the rich language and speech space to express uncertainties as more than mere numbers. |
| Researcher Affiliation | Academia | 1University of T ubingen. Now at Apple. 2Technical University of Munich.. |
| Pseudocode | No | The paper does not contain any structured pseudocode or algorithm blocks. It discusses theoretical concepts and proposes new research directions. |
| Open Source Code | No | The paper is a position paper and does not describe a methodology for which source code would typically be released. There is no mention of open-source code or code availability. |
| Open Datasets | No | The paper discusses research findings that sometimes refer to datasets like 'Natural Questions' (Kwiatkowski et al., 2019a) or 'Image Net-1k' (Mucs anyi et al., 2024), but it does not present its own experimental results based on these or other datasets, nor does it provide access information for any dataset it directly uses or releases. |
| Dataset Splits | No | The paper is theoretical and does not conduct experiments or analyze data, therefore, no dataset split information is provided. |
| Hardware Specification | No | The paper is theoretical and does not involve experimental runs requiring specific hardware. Thus, no hardware specifications are mentioned. |
| Software Dependencies | No | The paper is theoretical and does not describe any implementation details that would require specific software dependencies with version numbers. |
| Experiment Setup | No | The paper is a position paper focusing on theoretical concepts and new research avenues, not on empirical experimentation. Therefore, no experimental setup details, hyperparameters, or training configurations are provided. |