Temperature Scaling Attack Disrupting Model Confidence in Federated Learning
Not on the map. Why?
Abstract (arXiv)
Predictive confidence serves as a foundational control signal in mission-critical systems, directly governing risk-aware logic such as escalation, abstention, and conservative fallback. While prior federated learning attacks predominantly target accuracy or implant backdoors, we identify confidence calibration as a distinct attack objective. We present the Temperature Scaling Attack (TSA), a training-time attack that degrades calibration while preserving accuracy. By injecting temperature scaling with learning rate-temperature coupling during local training, TSA shifts model confidence while keeping predictive accuracy and common optimization signals close to benign training. We provide a convergence analysis under non-IID settings, showing that the coupling controls the primary update scale while leaving a bounded temperature-induced residual, yielding the standard non-convex FL convergence structure with an additional residual term. Across three benchmarks, TSA substantially shifts calibration (e.g., 145% error increase on CIFAR-100) with <2% accuracy change, and remains effective under robust aggregation and post-hoc calibration defenses. Case studies further show up to a 7.2x increase in missed verifications in healthcare and severe confidence-gating failures in autonomous driving, even when accuracy is unchanged. Overall, our results establish calibration integrity as a critical attack surface in federated learning.
Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 39 of its 64 references that are in the map. First posted on arXiv 6 Feb 2026.
Builds on
39 of 64 references.
References · 39
- 2015Distilling the Knowledge in a Neural Networkcited by 10,324
- 2016Communication-Efficient Learning of Deep Networks from Decentralized Datacited by 7,669
- 2020End-to-End Object Detection with Transformerscited by 7,414
- 2014Explaining and Harnessing Adversarial Examplescited by 7,364
- 2017Towards Deep Learning Models Resistant to Adversarial Attackscited by 6,415
- 2017On Calibration of Modern Neural Networkscited by 4,960
- 2016Pointer Sentinel Mixture Modelscited by 3,096
- 2016Towards Evaluating the Robustness of Neural Networkscited by 2,998
- 2019The Curious Case of Neural Text Degenerationcited by 2,916
- 2016Categorical Reparameterization with Gumbel-Softmaxcited by 2,739
- 2017What Uncertainties Do We Need in Bayesian Deep Learning for Computer Vision?cited by 2,352
- 2016Adversarial examples in the physical worldcited by 1,769
27 more
- 2017Selective Classification for Deep Neural Networkscited by 852
- 2019Measuring the Effects of Non-Identical Data Distribution for Federated Visual Classificationcited by 851
- 2018How To Backdoor Federated Learningcited by 765
- 2018Byzantine-Robust Distributed Learning: Towards Optimal Statistical Ratescited by 711
- 2019Local Model Poisoning Attacks to Byzantine-Robust Federated Learningcited by 493
- 2018Analyzing Federated Learning through an Adversarial Lenscited by 351
- 2020FLTrust: Byzantine-robust Federated Learning via Trust Bootstrappingcited by 346
- 2020Privacy and Robustness in Federated Learning: Attacks and Defensescited by 166
- 2018Mitigating Sybils in Federated Learning Poisoningcited by 161
- 2022FLDetector: Defending Federated Learning Against Model Poisoning Attacks via Detecting Malicious Clientscited by 125
- 2024EDT: Improving Large Language Models' Generation by Entropy-based Dynamic Temperature Samplingcited by 45
- 2023Rethinking Confidence Calibration for Failure Predictioncited by 43
- 2023Hot or Cold? Adaptive Temperature Sampling for Code Generation with Large Language Modelscited by 38
- 2020Temperature check: theory and practice for training models with softmax-cross-entropy lossescited by 31
- 2021Disrupting Deep Uncertainty Estimation Without Harming Accuracycited by 19
- 2022Curved Representation Space of Vision Transformerscited by 8
- 2023Task-Aware Risk Estimation of Perception Failures for Autonomous Vehiclescited by 7
- 2024Calibration Attacks: A Comprehensive Study of Adversarial Attacks on Model Confidencecited by 7
- 2024FedCal: Achieving Local and Global Calibration in Federated Learning via Aggregated Parameterized Scalercited by 6
- 2024Improving Local Training in Federated Learning via Temperature Scalingcited by 5
- 2024Invisible Optical Adversarial Stripes on Traffic Sign against Autonomous Vehiclescited by 5
- 2024DeTrigger: A Gradient-Centric Approach to Backdoor Attack Mitigation in Federated Learningcited by 5
- 2023Adversarial Attacks Against Uncertainty Quantificationcited by 4
- 2026IROS: A Dual-Process Architecture for Real-Time VLM-Based Indoor Navigationcited by 4
- 2023SureFED: Robust Federated Learning via Uncertainty-Aware Inward and Outward Inspectioncited by 3
- 2024Watch Your Head: Assembling Projection Heads to Save the Reliability of Federated Modelscited by 3
- 2024Tazza: Shuffling Neural Network Parameters for Secure and Private Federated Learningcited by 2
Built on by
Not cited yet.
Similar, not cited
Close in wording, no citation either way
- 2024Mitigating Malicious Attacks in Federated Learning via Confidence-aware Defense
- 2026FAROS: Robust Federated Learning with Adaptive Scaling against Backdoor Attacks
- 2025SoK: Benchmarking Poisoning Attacks and Defenses in Federated Learning
- 2025FedBAP: Backdoor Defense via Benign Adversarial Perturbation in Federated Learning
- 2026Revisiting Backdoor Threat in Federated Instruction Tuning from a Signal Aggregation Perspective
- 2025Evaluating Adversarial Attacks on Federated Learning for Temperature Forecasting
- 2026FedSurrogate: Backdoor Defense in Federated Learning via Layer Criticality and Surrogate Replacement
- 2025TrojanDam: Detection-Free Backdoor Defense in Federated Learning through Proactive Model Robustification utilizing OOD Data