Related papers: How Query Distribution Knowledge Breaks Multidimen…
When the training and test data are from different distributions, domain adaptation is needed to reduce dataset bias to improve the model's generalization ability. Since it is difficult to directly match the cross-domain joint…
In this work we derive a number of chain rules for mutual information quantities, suitable for analyzing quantum cryptography with imperfect devices that leak additional information to an adversary. First, we derive a chain rule between…
Query routing, the task to route user queries to different large language model (LLM) endpoints, can be considered as a text classification problem. However, out-of-distribution queries must be handled properly, as those could be about…
We introduce an explicit construction for a key distribution protocol in the Quantum Computational Timelock (QCT) security model, where one assumes that computationally secure encryption may only be broken after a time much longer than the…
Modern lattice-based cryptography, particularly the learning with errors paradigm, relies on injecting artificial noise to secure data against quantum adversaries. This study systematically examines the theoretical and physical boundaries…
Modern enterprise retrieval systems must handle short, underspecified queries such as ``foreign transaction fee refund'' and ``recent check status''. In these cases, semantic nuance and metadata matter but per-query large language model…
Large multiple-input multiple-output (MIMO) appears in massive multi-user MIMO and randomly-spread code-division multiple access (CDMA)-based wireless systems. In order to cope with the excessively high complexity of optimal data detection…
The Distributional Principal Autoencoder (DPA) combines distributionally correct reconstruction with principal-component-like interpretability of the encodings. In this work, we provide exact theoretical guarantees on both fronts. First, we…
We propose two quantum key distribution (QKD) protocols based on Bell's inequality, which can be considered as modified time-reversed E91 protocol. Similar to the measurement-device-independent quantum key distribution (MDI-QKD) protocol,…
Quantum key distribution (QKD) enables secure key sharing between distant parties, with several protocols proven resilient against conventional eavesdropping strategies. Here, we introduce a new attack scenario where an eavesdropper, Eve,…
Recently, it has been shown that Machine Learning models can leak sensitive information about their training data. This information leakage is exposed through membership and attribute inference attacks. Although many attack strategies have…
Quantum key distribution is the most well-known application of quantum cryptography. Previous proposed proofs of security of quantum key distribution contain various technical subtleties. Here, a conceptually simpler proof of security of…
Training models on a carefully chosen portion of data rather than the full dataset is now a standard preprocess for modern ML. From vision coreset selection to large-scale filtering in language models, it enables scalability with minimal…
Data lakes enable the training of powerful machine learning models on sensitive, high-value medical datasets, but also introduce serious privacy risks due to potential leakage of protected health information. Recent studies show adversaries…
A big challenge in continuous variable quantum key distribution is to prove security against arbitrary coherent attacks including realistic assumptions such as finite-size effects. Recently, such a proof has been presented in [Phys. Rev.…
Measurement-device-independent quantum key distribution (MDI-QKD) can remove all detection side-channels from quantum communication systems. The security proofs require, however, that certain assumptions on the sources are satisfied. This…
To effectively deploy Large Language Models (LLMs) in application-specific settings, fine-tuning techniques are applied to enhance performance on specialized tasks. This process often involves fine-tuning on user data data, which may…
The widespread use of publicly available datasets for training machine learning models raises significant concerns about data misuse. Availability attacks have emerged as a means for data owners to safeguard their data by designing…
Recent advancements in large language models (LLMs) have transformed natural language understanding and generation, leading to extensive benchmarking across diverse tasks. However, cryptanalysis - a critical area for data security and its…
A frequency permutation array (FPA) of length $n=m\lambda$ and distance $d$ is a set of permutations on a multiset over $m$ symbols, where each symbol appears exactly $\lambda$ times and the distance between any two elements in the array is…