Black-box machine learning models are now routinely used in high-risk settings, like
medical diagnostics, which demand uncertainty quantification to avoid consequential model …

Black-box machine learning models are now routinely used in high-risk settings, like
medical diagnostics, which demand uncertainty quantification to avoid consequential model …

Language models generate text based on successively sampling the next word. A decoding
procedure based on nucleus (top-$ p $) sampling chooses from the smallest possible set of …

Quantifying uncertainty in automatically generated text is important for letting humans check
potential hallucinations and making systems more reliable. Conformal prediction is an …

Conformal prediction (CP), a distribution-free uncertainty quantification (UQ) framework,
reliably provides valid predictive inference for black-box models. CP constructs prediction …

We study the performance of risk-controlling prediction sets (RCPS), an empirical risk
minimization-based formulation of conformal prediction, with a single trajectory of temporally …

Out-of-distribution (OOD) generalization has attracted increasing research attention in recent
years, due to its promising experimental results in real-world applications. In this paper, we …

The last decade in deep learning has brought on increasingly capable systems that are
deployed on a wide variety of applications. In natural language processing, the field has …

We study the split Conformal Prediction method when applied to Markovian data. We
quantify the gap in terms of coverage induced by the correlations in the data (compared to …

The convergence rate of a Markov chain to its stationary distribution is typically assessed
using the concept of total variation mixing time. However, this worst-case measure often …