Metraipolis › archive › dc_9bb16868a352
What Gradients Add to Text Leakage in Split Language Models, Counted per Token and per Document — arXiv:2610.04128
- Source
- arXiv
- Original
- https://arxiv.org/abs/2610.04128
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-06T05:28:41.966Z
- SHA-256 of the archived text
63f2d4cb707950f51442ac1982df9da34485cd14592118bc6e813ffd103a1a0c- Stable URL of this page
https://metraipolis.com/documento/dc_9bb16868a352
How to cite
What Gradients Add to Text Leakage in Split Language Models, Counted per Token and per Document — arXiv:2610.04128 — arXiv, https://arxiv.org/abs/2610.04128 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_9bb16868a352 (Metraipolis archive dc_9bb16868a352, sha256 63f2d4cb707950f51442ac1982df9da34485cd14592118bc6e813ffd103a1a0c)
BibTeX
@misc{dc9bb16868a352,
title = {What Gradients Add to Text Leakage in Split Language Models, Counted per Token and per Document — arXiv:2610.04128},
howpublished = {arXiv},
url = {https://arxiv.org/abs/2610.04128},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_9bb16868a352; sha256 63f2d4cb707950f51442ac1982df9da34485cd14592118bc6e813ffd103a1a0c},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv, https://arxiv.org/abs/2610.04128. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Split learning lets a client train a language model on a server without sending its text. The client runs the first layers itself and sends the server only their output, a vector of numbers for each token. During training, the server sends gradients back. We show that an observer at the split can rebuild most of the client's text from this traffic, and we measure how much the gradients help. On GPT-2, an attacker who holds only the publicly released weights of the client's layers recovers 94.20% of tokens from the activations alone and 97.38% when it also sees the gradients, 3.17 percentage points more 95% interval [2.72, 3.64]. Counted by document, the difference is much larger. The attacker rebuilds 13.71% of 32-token documents exactly without the gradients and 37.77% with them, because a document only counts when every token is right. How we count also changes how good a defence looks. Secret mixup, which blends each outgoing vector with a decoy, stops the attacker from rebuilding almost any document exactly, yet the attacker still recovers 83-91% of tokens. In a second experiment on GPT-2 and Qwen3-0.6B, where the server trains only a run of consecutive layers, the layer at which the run starts changes both model quality and leakage, even when the run's length is fixed. We recommend reporting leakage both per token and per document, and treating what a split model sends as being as sensitive as the text itself. Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.CL, nuovi lavori del giorno). Categorie arXiv: Cryptography and Security (cs.CR) ; Computation and Language (cs.CL); Machine Learning (cs.LG). Pagina del lavoro: https://arxiv.org/abs/2610.04128 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-10-06T05:24:20.400Z dalla pagina https://arxiv.org/list/cs.CL/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).