Pith. sign in

REVIEW 2 cited by

Improving Predictor Reliability with Selective Recalibration

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2410.05407 v1 pith:6O36NSLV submitted 2024-10-07 cs.LG cs.AI

classification cs.LGcs.AI
keywords recalibrationmodelselectivespacecalibrationconfidencedatadifficult
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

A reliable deep learning system should be able to accurately express its confidence with respect to its predictions, a quality known as calibration. One of the most effective ways to produce reliable confidence estimates with a pre-trained model is by applying a post-hoc recalibration method. Popular recalibration methods like temperature scaling are typically fit on a small amount of data and work in the model's output space, as opposed to the more expressive feature embedding space, and thus usually have only one or a handful of parameters. However, the target distribution to which they are applied is often complex and difficult to fit well with such a function. To this end we propose \textit{selective recalibration}, where a selection model learns to reject some user-chosen proportion of the data in order to allow the recalibrator to focus on regions of the input space that can be well-captured by such a model. We provide theoretical analysis to motivate our algorithm, and test our method through comprehensive experiments on difficult medical imaging and zero-shot classification tasks. Our results show that selective recalibration consistently leads to significantly lower calibration error than a wide range of selection and recalibration baselines.

Discussion (0). Sign in to comment.

Forward citations

Cited by 2 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Unsupervised Confidence Calibration for Reasoning LLMs from a Single Generation

    cs.LG 2026-04 unverdicted novelty 6.0 of 10

    Unsupervised single-generation confidence calibration for reasoning LLMs via offline self-consistency proxy distillation outperforms baselines on math and QA tasks and improves selective prediction.

  2. Identifying and Calibrating Overconfidence in Noisy Speech Recognition

    eess.AS 2025-09 conditional novelty 6.0 of 10

    In noisy speech, Whisper often assigns high confidence to wrong tokens; a selective token-level temperature-scaling calibrator reduces ECE by 58% on the R-SPIN low-SNR range.

Pith tools