works on

From the 2 of 6 linked papers with an AI index.

collaborators

6 papers

cs.LG2026

Privacy Leakage in Federated Learning in Radiology Reports: A Comparative Evaluation of Tokenizer-Driven Privacy Risks

Santhosh Parampottupadam, Andres Martinez, Dimitrios Bounias +3

The paper measures how much patient information can be reconstructed from model gradients in federated learning of radiology reports, comparing three different tokenizers and showi…

cs.LG2026

Inclusive Federated Learning Through Compliance-Weighted Noise Allocation in Healthcare AI

Santhosh Parampottupadam, Melih Coşğun, Sarthak Pati +7

The paper proposes a federated learning framework that adjusts differential privacy noise based on each healthcare institution's compliance level, allowing lower‑compliance sites t…

cs.CR2026

HADES: Privacy-Preserving Federated Learning via Selective Feature Encryption and Hybrid Model Fusion

Ergün Batuhan Kaynak, Kerem Bayramoglu, Sinem Sav

In this paper, we address the challenge of privacy-preserving training in federated learning (FL) by introducing a novel framework that selectively encrypts only the most privacy-s…

cs.CR2025

Bridging Local and Federated Data Normalization in Federated Learning: A Privacy-Preserving Approach

Melih Coşğun, Mert Gençtürk, Sinem Sav

Data normalization is a crucial preprocessing step for enhancing model performance and training stability. In federated learning (FL), where data remains distributed across multipl…

cs.CR2025

A Taxonomy of Attacks and Defenses in Split Learning

Aqsa Shabbir, Halil İbrahim Kanpak, Alptekin Küpçü +1

Split Learning (SL) has emerged as a promising paradigm for distributed deep learning, allowing resource-constrained clients to offload portions of their model computation to serve…

cs.CR2025

Generated Data with Fake Privacy: Hidden Dangers of Fine-tuning Large Language Models on Generated Data

Atilla Akkus, Masoud Poorghaffar Aghdam, Mingjie Li +4

Large language models (LLMs) have demonstrated significant success in various domain-specific tasks, with their performance often improving substantially after fine-tuning. However…