Salvato in:
Dettagli Bibliografici
Autori principali: Wang, Peiqi, Lam, Barbara D., Liu, Yingcheng, Asgari-Targhi, Ameneh, Panda, Rameswar, Wells, William M., Kapur, Tina, Golland, Polina
Natura: Preprint
Pubblicazione: 2024
Soggetti:
Accesso online:https://arxiv.org/abs/2410.04315
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866912304658907136
author Wang, Peiqi
Lam, Barbara D.
Liu, Yingcheng
Asgari-Targhi, Ameneh
Panda, Rameswar
Wells, William M.
Kapur, Tina
Golland, Polina
author_facet Wang, Peiqi
Lam, Barbara D.
Liu, Yingcheng
Asgari-Targhi, Ameneh
Panda, Rameswar
Wells, William M.
Kapur, Tina
Golland, Polina
contents We present a novel approach to calibrating linguistic expressions of certainty, e.g., "Maybe" and "Likely". Unlike prior work that assigns a single score to each certainty phrase, we model uncertainty as distributions over the simplex to capture their semantics more accurately. To accommodate this new representation of certainty, we generalize existing measures of miscalibration and introduce a novel post-hoc calibration method. Leveraging these tools, we analyze the calibration of both humans (e.g., radiologists) and computational models (e.g., language models) and provide interpretable suggestions to improve their calibration.
format Preprint
id arxiv_https___arxiv_org_abs_2410_04315
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Calibrating Expressions of Certainty
Wang, Peiqi
Lam, Barbara D.
Liu, Yingcheng
Asgari-Targhi, Ameneh
Panda, Rameswar
Wells, William M.
Kapur, Tina
Golland, Polina
Computation and Language
Machine Learning
We present a novel approach to calibrating linguistic expressions of certainty, e.g., "Maybe" and "Likely". Unlike prior work that assigns a single score to each certainty phrase, we model uncertainty as distributions over the simplex to capture their semantics more accurately. To accommodate this new representation of certainty, we generalize existing measures of miscalibration and introduce a novel post-hoc calibration method. Leveraging these tools, we analyze the calibration of both humans (e.g., radiologists) and computational models (e.g., language models) and provide interpretable suggestions to improve their calibration.
title Calibrating Expressions of Certainty
topic Computation and Language
Machine Learning
url https://arxiv.org/abs/2410.04315