AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

Target-Agnostic Calibration under Distribution Shift with Frequency-Aware Gradient Rectification

Yilin Zhang, Cai Xu, You Wu, Ziyu Guan, Wei Zhao

ICML 2026 · Label-efficient and robust learning › Recognition and long-tail · approach: Confidence calibration objectives

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add low-pass filtering without gradient rectification (Filter Only) (ablation)FGR · CIFAR-10 IDECE (%): 0.65 -> 11.54Hurt
add low-pass filtering to rectification (vs Rect. Only) (ablation)FGR · CIFAR-10-CECE (%): 18.60 -> 6.78Improved
add gradient rectification to filtering (vs Filter Only) (ablation)FGR · CIFAR-10 IDECE (%): 11.54 -> 0.65Improved
replace one-sided projection with weighted loss sum (ablation)FGR · CIFAR-100-CECE (%): 10.50 -> 11.80Hurt
use one-sided rectification instead of PCGrad (ablation)FGR · Camelyon17 testECE (%): 3.38 -> 1.14Improved
use one-sided rectification instead of CAGrad (ablation)FGR · Camelyon17 testECE (%): 1.28 -> 1.14Improved
use PCGrad instead of FGR rectification (ablation)FGR · iWildCam testaccuracy (%): 75.22 -> 77.21Improved
start FGR too early in training (ablation)FGR · CIFAR-10/100ID accuracyHurt
increase filtering ratio rho (ablation)FGR · CIFAR-10/100 and iWildCamshifted calibration vs clean performanceMixed
vary DCT strength lambda within {15,18,25} (ablation)FGR · CIFAR-10/100 and iWildCamcalibrationNo clear effect
two-stage fine-tuning (frozen backbone) instead of from-scratch FGR (ablation)FGR · CIFAR-10 ID (ResNet-50)ECE (%): 0.65 -> 0.58Improved
apply FGR to transformer backbones instead of CESwin-Tiny · Camelyon17-OODECE (%): 15.96 -> 8.21Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 0 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy