Automating chest radiograph interpretation using Deep Learning (DL) models has the potential to significantly improve clinical workflows, decision-making, and large-scale health screening. However, in medical settings, merely optimising predictive performance is insufficient, as the quantification of uncertainty is equally crucial. This paper investigates the relationship between predictive uncertainty, derived from Bayesian Deep Learning approximations, and human/linguistic uncertainty, as estimated from free-text radiology reports labelled by rule-based labellers. Utilising BERT as the model of choice, this study evaluates different binarisation methods for uncertainty labels and explores the efficacy of Monte Carlo Dropout and Deep Ensembles in estimating predictive uncertainty. The results demonstrate good model performance, but also a modest correlation between predictive and linguistic uncertainty, highlighting the challenges in aligning machine uncertainty with human interpretation nuances. Our findings suggest that while Bayesian approximations provide valuable uncertainty estimates, further refinement is necessary to fully capture and utilise the subtleties of human uncertainty in clinical applications.

Learning human preferences is essential for human-robot interaction, as it enables robots to adapt their behaviors to align with human expectations and goals. However, the inherent uncertainties in both human behavior and robotic systems make preference learning a challenging task. While probabilistic robotics algorithms offer uncertainty quantification, the integration of human preference uncertainty remains underexplored. To bridge this gap, we introduce uncertainty unification and propose a novel framework, uncertainty-unified preference learning (UUPL), which enhances Gaussian Process (GP)-based preference learning by unifying human and robot uncertainties. Specifically, UUPL includes a human preference uncertainty model that improves GP posterior mean estimation, and an uncertainty-weighted Gaussian Mixture Model (GMM) that enhances GP predictive variance accuracy. Additionally, we design a user-specific calibration process to align uncertainty representations across users, ensuring consistency and reliability in the model performance. Comprehensive experiments and user studies demonstrate that UUPL achieves state-of-the-art performance in both prediction accuracy and user rating. An ablation study further validates the effectiveness of human uncertainty model and uncertainty-weighted GMM of UUPL.

Recent work has sought to quantify large language model uncertainty to facilitate model control and modulate user trust. Previous works focus on measures of uncertainty that are theoretically grounded or reflect the average overt behavior of the model. In this work, we investigate a variety of uncertainty measures, in order to identify measures that correlate with human group-level uncertainty. We find that Bayesian measures and a variation on entropy measures, top-k entropy, tend to agree with human behavior as a function of model size. We find that some strong measures decrease in human-similarity with model size, but, by multiple linear regression, we find that combining multiple uncertainty measures provide comparable human-alignment with reduced size-dependency.

The effectiveness of automatic evaluation of generative models is typically measured by comparing it to human evaluation using correlation metrics. However, metrics like Krippendorff's $\alpha$ and Randolph's $\kappa$, originally designed to measure the reliability of human labeling, make assumptions about human behavior and the labeling process. In this paper, we show how *relying on a single aggregate correlation score* can obscure fundamental differences between human behavior and automatic evaluation methods, including LLM-as-a-Judge. Specifically, we demonstrate that when the proportion of samples with variation or uncertainty in human labels (gathered during human evaluation) is relatively high, machine labels (generated by automatic evaluation methods) may superficially appear to have similar or better correlation with the human majority label compared to human-to-human (HH) correlation. This can create the illusion that automatic evaluation approximates the human majority label. However, as the proportion of samples with consistent human labels increases, the correlation between machine and human labels fall well below HH correlation. Based on these findings, we first propose stratifying results by human label uncertainty to provide a more robust analysis of automatic evaluation performance. Second, recognizing that uncertainty and variation are inherent in perception-based human evaluations, such as those involving attitudes or preferences, we introduce a new metric - *binned Jensen-Shannon Divergence for perception* for such scenarios to better measure the effectiveness of automatic evaluations. Third, we present visualization techniques -- *perception charts*, to compare the strengths and limitations of automatic evaluation and to contextualize correlation measures appropriately

Large vision-language models frequently struggle to accurately predict responses provided by multiple human annotators, particularly when those responses exhibit human uncertainty. In this study, we focus on the Visual Question Answering (VQA) task, and we comprehensively evaluate how well the state-of-the-art vision-language models correlate with the distribution of human responses. To do so, we categorize our samples based on their levels (low, medium, high) of human uncertainty in disagreement (HUD) and employ not only accuracy but also three new human-correlated metrics in VQA, to investigate the impact of HUD. To better align models with humans, we also verify the effect of common calibration and human calibration. Our results show that even BEiT3, currently the best model for this task, struggles to capture the multi-label distribution inherent in diverse human responses. Additionally, we observe that the commonly used accuracy-oriented calibration technique adversely affects BEiT3's ability to capture HUD, further widening the gap between model predictions and human distributions. In contrast, we show the benefits of calibrating models towards human distributions for VQA, better aligning model confidence with human uncertainty. Our findings highlight that for VQA, the consistent alignment between human responses and model predictions is understudied and should become the next crucial target of future studies.

Redacted emails satisfy most privacy requirements but they make it more difficult to detect anomalous emails that may be indicative of data exfiltration. In this paper we develop an enhanced method of Active Learning using an information gain maximizing heuristic, and we evaluate its effectiveness in a real world setting where only redacted versions of email could be labeled by human analysts due to privacy concerns. In the first case study we examined how Active Learning should be carried out. We found that model performance was best when a single highly skilled (in terms of the labelling task) analyst provided the labels. In the second case study we used confidence ratings to estimate the labeling uncertainty of analysts and then prioritized instances for labeling based on the expected information gain (the difference between model uncertainty and analyst uncertainty) that would be provided by labelling each instance. We found that the information maximization gain heuristic improved model performance over existing sampling methods for Active Learning. Based on the results obtained, we recommend that analysts should be screened, and possibly trained, prior to implementation of Active Learning in cybersecurity applications. We also recommend that the information gain maximizing sample method (based on expert confidence) should be used in early stages of Active Learning, providing that well-calibrated confidence can be obtained. We also note that the expertise of analysts should be assessed prior to Active Learning, as we found that analysts with lower labelling skill had poorly calibrated (over-) confidence in their labels.

This paper proposes a new control algorithm for human-robot co-transportation based on a robot manipulator equipped with a mobile base and a robotic arm. The primary focus is to adapt to human uncertainties through the robot's whole-body dynamics and pose optimization. We introduce an augmented Model Predictive Control (MPC) formulation that explicitly models human uncertainties and contains extra variables than regular MPC to optimize the pose of the robotic arm. The core of our methodology involves a two-step iterative design: At each planning horizon, we select the best pose of the robotic arm (joint angle combination) from a candidate set, aiming to achieve the lowest estimated control cost. This selection is based on solving an uncertainty-aware Discrete Algebraic Ricatti Equation (DARE), which also informs the optimal control inputs for both the mobile base and the robotic arm. To validate the effectiveness of the proposed approach, we provide theoretical derivation for the uncertainty-aware DARE and perform simulated and proof-of-concept hardware experiments using a Fetch robot under varying conditions, including different nominal trajectories and noise levels. The results reveal that our proposed approach outperforms baseline algorithms, maintaining similar execution time with that do not consider human uncertainty or do not perform pose optimization.

Aligning model representations to humans has been found to improve robustness and generalization. However, such methods often focus on standard observational data. Synthetic data is proliferating and powering many advances in machine learning; yet, it is not always clear whether synthetic labels are perceptually aligned to humans -- rendering it likely model representations are not human aligned. We focus on the synthetic data used in mixup: a powerful regularizer shown to improve model robustness, generalization, and calibration. We design a comprehensive series of elicitation interfaces, which we release as HILL MixE Suite, and recruit 159 participants to provide perceptual judgments along with their uncertainties, over mixup examples. We find that human perceptions do not consistently align with the labels traditionally used for synthetic points, and begin to demonstrate the applicability of these findings to potentially increase the reliability of downstream models, particularly when incorporating human uncertainty. We release all elicited judgments in a new data hub we call H-Mix.

Placing a human in the loop may abate the risks of deploying AI systems in safety-critical settings (e.g., a clinician working with a medical AI system). However, mitigating risks arising from human error and uncertainty within such human-AI interactions is an important and understudied issue. In this work, we study human uncertainty in the context of concept-based models, a family of AI systems that enable human feedback via concept interventions where an expert intervenes on human-interpretable concepts relevant to the task. Prior work in this space often assumes that humans are oracles who are always certain and correct. Yet, real-world decision-making by humans is prone to occasional mistakes and uncertainty. We study how existing concept-based models deal with uncertain interventions from humans using two novel datasets: UMNIST, a visual dataset with controlled simulated uncertainty based on the MNIST dataset, and CUB-S, a relabeling of the popular CUB concept dataset with rich, densely-annotated soft labels from humans. We show that training with uncertain concept labels may help mitigate weaknesses of concept-based systems when handling uncertain interventions. These results allow us to identify several open challenges, which we argue can be tackled through future multidisciplinary research on building interactive uncertainty-aware systems. To facilitate further research, we release a new elicitation platform, UElic, to collect uncertain feedback from humans in collaborative prediction tasks.

Calibration is a popular framework to evaluate whether a classifier knows when it does not know - i.e., its predictive probabilities are a good indication of how likely a prediction is to be correct. Correctness is commonly estimated against the human majority class. Recently, calibration to human majority has been measured on tasks where humans inherently disagree about which class applies. We show that measuring calibration to human majority given inherent disagreements is theoretically problematic, demonstrate this empirically on the ChaosNLI dataset, and derive several instance-level measures of calibration that capture key statistical properties of human judgements - class frequency, ranking and entropy.