When Shuffling Fails: Token Recovery from Obfuscated KV Caches

LLM serving systems store, offload, and share KV caches, exposing token-level representations to infrastructure operators and co-tenants. Shuffled obfuscation combines secret linear transforms with fresh permutations, claiming security from b! possible shufflings. We show that input symmetry invalidates this counting argument: security depends on distinguishable outputs under chosen inputs. Repeated tokens produce identical first-layer Value rows, making every row permutation leave the data unchanged. Only auxiliary structures such as a single-row recovery marker vary, reducing the effective state space from b! to b. We construct Repeat-to-Reveal. The attacker collects the b distinguishable states from its own probes and exploits their rank-one differences to recover an equivalent denoiser without knowing the defender's secret parameters. A one-time calibration transfers across requests sharing the same secret epoch, enabling passive token-multiset recovery from subsequently observed cache blocks through public-model dictionary matching. We instantiate the attack on official KV-Cloak and evaluate 3,000 real WildChat user prompts across nine models and eleven configurations. The attack recovers the secret parameters with high accuracy and achieves 100 percent token-multiset recovery on GPU. Refreshing row-level secrets between epochs interrupts cross-request transfer, confirming secret reuse scope as the operative vulnerability. We further confirm orbit collapse on two other open-source defenses: the sequence-permutation implementation of Thomas et al. and PermLLM. The KV-Cloak developers invited a detailed report through their private vulnerability reporting channel.

Authors

Publication Details

Journal
arXiv (Cornell University)
Published
2026-11-18
DOI
https://doi.org/10.48550/arxiv.2611.24276
Primary Topic
Adversarial Robustness in Machine Learning
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

When Shuffling Fails: Token Recovery from Obfuscated KV Caches

Chengzhi Yi, Haoyu Wang, Wei Song, Xiao Cheng
arXiv (Cornell University)
Adversarial Robustness in Machine Learning
preprint

When Shuffling Fails: Token Recovery from Obfuscated KV Caches

Chengzhi Yi, Haoyu Wang, Wei Song, Xiao Cheng
preprint en

Abstract

LLM serving systems store, offload, and share KV caches, exposing token-level representations to infrastructure operators and co-tenants. Shuffled obfuscation combines secret linear transforms with fresh permutations, claiming security from b! possible shufflings. We show that input symmetry invalidates this counting argument: security depends on distinguishable outputs under chosen inputs. Repeated tokens produce identical first-layer Value rows, making every row permutation leave the data unchanged. Only auxiliary structures such as a single-row recovery marker vary, reducing the effective state space from b! to b. We construct Repeat-to-Reveal. The attacker collects the b distinguishable states from its own probes and exploits their rank-one differences to recover an equivalent denoiser without knowing the defender's secret parameters. A one-time calibration transfers across requests sharing the same secret epoch, enabling passive token-multiset recovery from subsequently observed cache blocks through public-model dictionary matching. We instantiate the attack on official KV-Cloak and evaluate 3,000 real WildChat user prompts across nine models and eleven configurations. The attack recovers the secret parameters with high accuracy and achieves 100 percent token-multiset recovery on GPU. Refreshing row-level secrets between epochs interrupts cross-request transfer, confirming secret reuse scope as the operative vulnerability. We further confirm orbit collapse on two other open-source defenses: the sequence-permutation implementation of Thomas et al. and PermLLM. The KV-Cloak developers invited a detailed report through their private vulnerability reporting channel.

arXiv (Cornell University)
Adversarial Robustness in Machine Learning
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon