Copertina

Abstract Gradient Training: A Unified Certification Framework for Data Poisoning, Unlearning, and Differential Privacy

Fuente: arXiv

Salvato in:

Dettagli Bibliografici
Autori principali:	Sosnin, Philip, Wicker, Matthew, Collyer, Josh, Tsay, Calvin
Natura:	Preprint
Pubblicazione:	2025
Soggetti:	Machine Learning
Accesso online:	Acceder al recurso
Tags:	Aggiungi Tag Nessun Tag, puoi essere il primo ad aggiungerne!!

Documenti analoghi

Exact Certification of Data-Poisoning Attacks Using Mixed-Integer Programming
di: Sosnin, Philip, et al.
Pubblicazione: (2026)

Certified Robustness to Data Poisoning in Gradient-Based Training
di: Sosnin, Philip, et al.
Pubblicazione: (2024)

Certification for Differentially Private Prediction in Gradient-Based Training
di: Wicker, Matthew, et al.
Pubblicazione: (2024)

Provably Safe Model Updates
di: Elmecker-Plakolm, Leo, et al.
Pubblicazione: (2025)

Towards Poisoning Robustness Certification for Natural Language Generation
di: Ghitu, Mihnea, et al.
Pubblicazione: (2026)

Scaling Mixed-Integer Programming for Certification of Neural Network Controllers Using Bounds Tightening
di: Sosnin, Philip, et al.
Pubblicazione: (2024)

Relaxation-Informed Training of Neural Network Surrogate Models
di: Tsay, Calvin
Pubblicazione: (2026)

Training Neural ODEs Using Fully Discretized Simultaneous Optimization
di: Shapovalova, Mariia, et al.
Pubblicazione: (2025)

Variational Routing: A Scalable Bayesian Framework for Calibrated Mixture-of-Experts Transformers
di: Li, Albus Yizhuo, et al.
Pubblicazione: (2026)

Bayesian optimization as a flexible and efficient design framework for sustainable process systems
di: Paulson, Joel A., et al.
Pubblicazione: (2024)

Poisoning Attacks to Local Differential Privacy Protocols for Trajectory Data
di: Hsu, I-Jung, et al.
Pubblicazione: (2025)

A Unified Gradient-based Framework for Task-agnostic Continual Learning-Unlearning
di: Huang, Zhehao, et al.
Pubblicazione: (2025)

P-split formulations: A class of intermediate formulations between big-M and convex hull for disjunctive constraints
di: Kronqvist, Jan, et al.
Pubblicazione: (2022)

Potion: Towards Poison Unlearning
di: Schoepf, Stefan, et al.
Pubblicazione: (2024)

Class Machine Unlearning for Complex Data via Concepts Inference and Data Poisoning
di: Chang, Wenhan, et al.
Pubblicazione: (2024)

Data Efficient Subset Training with Differential Privacy
di: Gandhi, Ninad Jayesh, et al.
Pubblicazione: (2025)

A Unified Framework for Continual Learning and Unlearning
di: Chatterjee, Romit, et al.
Pubblicazione: (2024)

When Deep Learning Meets Polyhedral Theory: A Survey
di: Huchette, Joey, et al.
Pubblicazione: (2023)

Vertical Federated Unlearning via Backdoor Certification
di: Han, Mengde, et al.
Pubblicazione: (2024)

Efficient Federated Unlearning with Adaptive Differential Privacy Preservation
di: Jiang, Yu, et al.
Pubblicazione: (2024)

Model Guidance via Robust Feature Attribution
di: Ghitu, Mihnea, et al.
Pubblicazione: (2025)

BlockRR: A Unified Framework of RR-type Algorithms for Label Differential Privacy
di: Liu, Haixia, et al.
Pubblicazione: (2026)

On Evaluating the Poisoning Robustness of Federated Learning under Local Differential Privacy
di: Wang, Zijian, et al.
Pubblicazione: (2025)

Less Noise, Same Certificate: Retain Sensitivity for Unlearning
di: Heinzler, Carolin, et al.
Pubblicazione: (2026)

Machine Unlearning Fails to Remove Data Poisoning Attacks
di: Pawelczyk, Martin, et al.
Pubblicazione: (2024)

Unified Gradient-Based Machine Unlearning with Remain Geometry Enhancement
di: Huang, Zhehao, et al.
Pubblicazione: (2024)

Targeted Unlearning with Single Layer Unlearning Gradient
di: Cai, Zikui, et al.
Pubblicazione: (2024)

Global Optimization of Gaussian Process Acquisition Functions Using a Piecewise-Linear Kernel Approximation
di: Xie, Yilin, et al.
Pubblicazione: (2024)

Modeling High-Dimensional Dependent Data in the Presence of Many Explanatory Variables and Weak Signals
di: Gao, Zhaoxing, et al.
Pubblicazione: (2024)

SafeAdapt: Provably Safe Policy Updates in Deep Reinforcement Learning
di: Anisimov, Maksim, et al.
Pubblicazione: (2026)

Inverting Gradient Attacks Makes Powerful Data Poisoning
di: Bouaziz, Wassim, et al.
Pubblicazione: (2024)

Hidden Poison: Machine Unlearning Enables Camouflaged Poisoning Attacks
di: Di, Jimmy Z., et al.
Pubblicazione: (2022)

Exact Certification of (Graph) Neural Networks Against Label Poisoning
di: Sabanayagam, Mahalakshmi, et al.
Pubblicazione: (2024)

Enhancing the Antidote: Improved Pointwise Certifications against Poisoning Attacks
di: Liu, Shijie, et al.
Pubblicazione: (2023)

Exact Certification of Neural Networks and Partition Aggregation Ensembles against Label Poisoning
di: Mohgaonkar, Ajinkya, et al.
Pubblicazione: (2026)

Poisoning Attacks and Defenses to Federated Unlearning
di: Wang, Wenbin, et al.
Pubblicazione: (2025)

Global optimization of graph acquisition functions for neural architecture search
di: Xie, Yilin, et al.
Pubblicazione: (2025)

An Efficient Spatial Branch-and-Bound Algorithm for Global Optimization of Gaussian Process Posterior Mean Functions
di: Tang, Wei-Ting, et al.
Pubblicazione: (2026)

EARL-BO: Reinforcement Learning for Multi-Step Lookahead, High-Dimensional Bayesian Optimization
di: Cheon, Mujin, et al.
Pubblicazione: (2024)

FROC: A Unified Framework with Risk-Optimized Control for Machine Unlearning in LLMs
di: Goh, Si Qi, et al.
Pubblicazione: (2025)