LightSplit uses non-invertible orthogonal projections as an information bottleneck in split learning to reduce transmitted dimensionality by 32x while retaining more than 95% accuracy and limiting reconstruction risk.
Feature space hijacking attacks against differentially private split learning
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
citation-role summary
background 1
citation-polarity summary
years
2026 2verdicts
UNVERDICTED 2roles
background 1polarities
background 1representative citing papers
A survey that introduces a unified training pipeline and taxonomizes split learning approaches for LLM fine-tuning across model, system, and privacy dimensions.
citing papers explorer
-
LightSplit: Practical Privacy-Preserving Split Learning via Orthogonal Projections
LightSplit uses non-invertible orthogonal projections as an information bottleneck in split learning to reduce transmitted dimensionality by 32x while retaining more than 95% accuracy and limiting reconstruction risk.
-
A Survey on Split Learning for LLM Fine-Tuning: Models, Systems, and Privacy Optimizations
A survey that introduces a unified training pipeline and taxonomizes split learning approaches for LLM fine-tuning across model, system, and privacy dimensions.