Deep ensembles fail to capture meaningful epistemic uncertainty in message-passing GNNs due to epistemic collapse where independently trained networks converge to similar predictions.
Title resolution pending
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
years
2026 2representative citing papers
Larger LLMs handle detailed crystal descriptions better than small ones, and mean negative log-likelihood of predicted numbers tracks prediction error after fine-tuning.
citing papers explorer
-
Do Deep Ensembles Actually Capture Uncertainty in Graph Neural Networks?
Deep ensembles fail to capture meaningful epistemic uncertainty in message-passing GNNs due to epistemic collapse where independently trained networks converge to similar predictions.
-
Scale-Dependent Input Representation and Confidence Estimation for LLMs in Materials Property Prediction
Larger LLMs handle detailed crystal descriptions better than small ones, and mean negative log-likelihood of predicted numbers tracks prediction error after fine-tuning.