Research map / Federated learning and privacy / Privacy attacks and secure learning
Stealing Split Learning Bottom Models by Recovering Embedding Geometry
Qinbo Zhang, Yanhang Shi, Ziyi Zhang, Hao Wang, Sai Qian Zhang, Jian Li
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 changed | Where | Measured | Effect |
|---|---|---|---|
| add neighbor matching in the contrastive space (ablation) | VENOM · CIFAR-10, Model Rake | S-ACC (%): 24.72 -> 52.58 | Improved |
| use contrastive space instead of raw embedding space for neighbor mining (ablation) | VENOM · CIFAR-10, Model Rake | S-ACC (%): 32.03 -> 52.58 | Improved |
| add the KL feature-shape term (ablation) | VENOM · CIFAR-10, Model Rake | S-ACC (%): 46.43 -> 52.58 | Improved |
| add the KL feature-shape term (ablation) | VENOM · CIFAR-10, Vanilla (no defense) | S-ACC (%): 68.24 -> 69.97 | Improved |
| increase neighbor set size k from 1% to 10% of auxiliary set (ablation) | VENOM · CIFAR-10 | S-ACC and AGR | Improved |
| increase neighbor size from 10% to 15% (ablation) | VENOM · CIFAR-10 | S-ACC and AGR | No clear effect |
| set attraction-repulsion weight alpha to 0.5 instead of skewed values (ablation) | VENOM · CIFAR-10 | S-ACC and AGR | Improved |
| increase auxiliary data from 1% to 10% of dataset (ablation) | VENOM · CIFAR-10 | S-ACC and AGR | Improved |
| use near-OOD auxiliary data (CIFAR-100, Tiny-ImageNet) instead of in-distribution (ablation) | VENOM · CIFAR-10 | S-ACC and AGR | Hurt |
| use far-OOD auxiliary data (MNIST) (ablation) | VENOM · CIFAR-10 | S-ACC and AGR | Hurt |
| deepen surrogate beyond the target capacity (3 blocks) (ablation) | VENOM · CIFAR-10 | S-ACC and AGR | No clear effect |
| use approximate subset-based neighbor selection instead of exact search (ablation) | VENOM · SUSY, Model Rake | S-ACC (%): 68.81 -> 63.16 | Hurt |
Closest papers by meaning
- Constructing Adversarial Examples for Vertical Federated Learning: Optimal Client Corruption through Multi-Armed Bandit (ICLR 2024 · similarity 0.85)
- DataStealing: Steal Data from Diffusion Models in Federated Learning with Multiple Trojans (NeurIPS 2024 · similarity 0.84)
- FedNE: Surrogate-Assisted Federated Neighbor Embedding for Dimensionality Reduction (NeurIPS 2024 · similarity 0.84)
- Concealing Sensitive Samples against Gradient Leakage in Federated Learning (AAAI 2024 · similarity 0.83)
- GRAIN: Exact Graph Reconstruction from Gradients (ICLR 2025 · similarity 0.83)
- Detecting Backdoor Attacks in Federated Learning via Direction Alignment Inspection (CVPR 2025 · similarity 0.83)
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.
