Integrating over multiple forms of statistical uncertainty associated with serological surveys can improve serosurvey design while also enabling that uncertainty to be appropriately propagated through epidemiological models.

A new machine learning method makes it possible to efficiently identify the parameters of cognitive models using Bayesian inference, even when only model simulations are available.

Quantification of uncertainty in drug dose-responses provides a more robust metric for differentiating drugs and identifying biomarkers of response.

µGUIDE is a Bayesian framework that leverages simulation-based inference to efficiently estimate posterior distributions of any forward model parameters, allowing for uncertainty quantification and degeneracy detection.

Analysis of performance on a novel behavioural paradigm investigating active information gathering reveals that subjective cognitive impairment is associated with hyperreactivity to uncertainty and links this mechanism to the affective burden in the condition and heightened insular-hippocampal connectivity.

A mathematical model of bias in marker-gene and metagenomic sequencing measurements explains systematic errors in defined mixtures of microbial species, and enables quantitative and reproducible investigation of biological communities.

For perceptual inference, human observers do not estimate sensory uncertainty instantaneously from the current sensory signals alone, but by combining past and current sensory inputs consistent with a Bayesian learner.

Distortion and elimination of limb visual feedback affects low-level stretch reflex control, indicating the involvement of a high-level and multimodal representation of the limb state in orchestrating hierarchical sensorimotor control.

Increased usability and validity of neuroscience models, through FAIR workflows for the whole modeling process, including data and model management, parameter estimation, uncertainty quantification, and model analysis.

The rodent brain represents uncertainty associated with short-term predictions during naturalistic navigation tasks sequentially by sampling hypothetical future trajectories in every ~100 ms, corresponding to successive theta cycles.