A statistically consistent measure of semantic uncertainty using Language Models
This provides a method for researchers and practitioners to better assess uncertainty in language model predictions, though it is incremental as it builds on existing uncertainty quantification techniques.
The authors tackled the problem of quantifying uncertainty in language model outputs by proposing a statistically consistent measure called semantic spectral entropy, which they demonstrated through simulations to be accurate and robust even with inherent randomness.
To address the challenge of quantifying uncertainty in the outputs generated by language models, we propose a novel measure of semantic uncertainty, semantic spectral entropy, that is statistically consistent under mild assumptions. This measure is implemented through a straightforward algorithm that relies solely on standard, pretrained language models, without requiring access to the internal generation process. Our approach imposes minimal constraints on the choice of language models, making it broadly applicable across different architectures and settings. Through comprehensive simulation studies, we demonstrate that the proposed method yields an accurate and robust estimate of semantic uncertainty, even in the presence of the inherent randomness characteristic of generative language model outputs.