Related papers: Predicting generalization performance with correctness discriminators

Predicting generalization performance with correctness discriminators

URL: http://arxiv.org/abs/2311.09422v1
Date: Wed, 15 Nov 2023 22:43:42 GMT
Title: Predicting generalization performance with correctness discriminators
Authors: Yuekun Yao and Alexander Koller
Abstract summary: We present a novel model that establishes upper and lower bounds on the accuracy, without requiring gold labels for the unseen data. We show across a variety of tagging, parsing, and semantic parsing tasks that the gold accuracy is reliably between the predicted upper and lower bounds.
Score: 64.00420578048855
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Abstract: The ability to predict an NLP model's accuracy on unseen, potentially out-of-distribution data is a prerequisite for trustworthiness. We present a novel model that establishes upper and lower bounds on the accuracy, without requiring gold labels for the unseen data. We achieve this by training a discriminator which predicts whether the output of a given sequence-to-sequence model is correct or not. We show across a variety of tagging, parsing, and semantic parsing tasks that the gold accuracy is reliably between the predicted upper and lower bounds, and that these bounds are remarkably close together.

Related papers

Provably Reliable Conformal Prediction Sets in the Presence of Data Poisoning [53.42244686183879]
Conformal prediction provides model-agnostic and distribution-free uncertainty quantification. Yet, conformal prediction is not reliable under poisoning attacks where adversaries manipulate both training and calibration data. We propose reliable prediction sets (RPS): the first efficient method for constructing conformal prediction sets with provable reliability guarantees under poisoning.
arXiv Detail & Related papers (2024-10-13T15:37:11Z)
Measuring Stochastic Data Complexity with Boltzmann Influence Functions [12.501336941823627]
Estimating uncertainty of a model's prediction on a test point is a crucial part of ensuring reliability and calibration under distribution shifts. We propose IF-COMP, a scalable and efficient approximation of the pNML distribution that linearizes the model with a temperature-scaled Boltzmann influence function. We experimentally validate IF-COMP on uncertainty calibration, mislabel detection, and OOD detection tasks, where it consistently matches or beats strong baseline methods.
arXiv Detail & Related papers (2024-06-04T20:01:39Z)
VisFIS: Visual Feature Importance Supervision with Right-for-the-Right-Reason Objectives [84.48039784446166]
We show that model FI supervision can meaningfully improve VQA model accuracy as well as performance on several Right-for-the-Right-Reason metrics. Our best performing method, Visual Feature Importance Supervision (VisFIS), outperforms strong baselines on benchmark VQA datasets. Predictions are more accurate when explanations are plausible and faithful, and not when they are plausible but not faithful.
arXiv Detail & Related papers (2022-06-22T17:02:01Z)
Approximate Conditional Coverage via Neural Model Approximations [0.030458514384586396]
We analyze a data-driven procedure for obtaining empirically reliable approximate conditional coverage. We demonstrate the potential for substantial (and otherwise unknowable) under-coverage with split-conformal alternatives with marginal coverage guarantees.
arXiv Detail & Related papers (2022-05-28T02:59:05Z)
Striking a Balance: Alleviating Inconsistency in Pre-trained Models for Symmetric Classification Tasks [4.971443651456398]
inconsistency is often observed in the predicted labels or confidence scores. We highlight this model shortcoming and apply a consistency loss function to alleviate inconsistency in symmetric classification. Our results show an improved consistency in predictions for three paraphrase detection datasets without a significant drop in the accuracy scores.
arXiv Detail & Related papers (2022-03-25T07:55:39Z)
Leveraging Unlabeled Data to Predict Out-of-Distribution Performance [63.740181251997306]
Real-world machine learning deployments are characterized by mismatches between the source (training) and target (test) distributions. In this work, we investigate methods for predicting the target domain accuracy using only labeled source data and unlabeled target data. We propose Average Thresholded Confidence (ATC), a practical method that learns a threshold on the model's confidence, predicting accuracy as the fraction of unlabeled examples.
arXiv Detail & Related papers (2022-01-11T23:01:12Z)
Distribution-free uncertainty quantification for classification under label shift [105.27463615756733]
We focus on uncertainty quantification (UQ) for classification problems via two avenues. We first argue that label shift hurts UQ, by showing degradation in coverage and calibration. We examine these techniques theoretically in a distribution-free framework and demonstrate their excellent practical performance.
arXiv Detail & Related papers (2021-03-04T20:51:03Z)
Unlabelled Data Improves Bayesian Uncertainty Calibration under Covariate Shift [100.52588638477862]
We develop an approximate Bayesian inference scheme based on posterior regularisation. We demonstrate the utility of our method in the context of transferring prognostic models of prostate cancer across globally diverse populations.
arXiv Detail & Related papers (2020-06-26T13:50:19Z)
Estimation of Accurate and Calibrated Uncertainties in Deterministic models [0.8702432681310401]
We devise a method to transform a deterministic prediction into a probabilistic one. We show that for doing so, one has to compromise between the accuracy and the reliability (calibration) of such a model. We show several examples both with synthetic data, where the underlying hidden noise can accurately be recovered, and with large real-world datasets.
arXiv Detail & Related papers (2020-03-11T04:02:56Z)

This list is automatically generated from the titles and abstracts of the papers in this site.