RevertLM reconstructs private text from intermediate LLM representations in split learning using a projection into embedding space plus a generative decoder, outperforming prior embedding-only attacks.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Model Inversion in Split Learning for Personalized LLMs: New Insights from Information Bottleneck Theory
RevertLM reconstructs private text from intermediate LLM representations in split learning using a projection into embedding space plus a generative decoder, outperforming prior embedding-only attacks.