38 citations · 85 across the 13 of their papers we have counts for
13 papers
KV Cache Compression Through the Lens of Transform Coding
Hannah Laus, Claudio Mayrink Verdun, Hao Wang +2
The key-value (KV) cache stores information from past tokens and is a major memory bottleneck in long-context inference. Existing quantization methods address this bottleneck by re…
Optimal conversion from Rényi Differential Privacy to -Differential Privacy
Anneliese Riess, Juan Felipe Gomez, Flavio du Pin Calmon +2
We prove the conjecture stated in Appendix F.3 of \citet{zhu2022optimalaccountingdifferentialprivacy}: among all conversion rules that map a Rényi Differential Privacy (RDP) profil…
AI Alignment at Your Discretion
Maarten Buyl, Hadi Khalaf, Claudio Mayrink Verdun +3
In AI alignment, extensive latitude must be granted to annotators, either human or algorithmic, to judge which model outputs are `better' or `safer.' We refer to this latitude as a…
Attack-Aware Noise Calibration for Differential Privacy
Bogdan Kulynych, Juan Felipe Gomez, Georgios Kaissis +2
Differential privacy (DP) is a widely used approach for mitigating privacy risks when training machine learning models on sensitive data. DP mechanisms add noise during training to…
Tunable Measures for Information Leakage and Applications to Privacy-Utility Tradeoffs
Jiachun Liao, Oliver Kosut, Lalitha Sankar +1
We introduce a tunable measure for information leakage called maximal alpha-leakage. This measure quantifies the maximal gain of an adversary in inferring any (potentially random)…
Hypothesis Testing under Mutual Information Privacy Constraints in the High Privacy Regime
Jiachun Liao, Lalitha Sankar, Vincent Y. F. Tan +1
Hypothesis testing is a statistical inference framework for determining the true distribution among a set of possible distributions for a given dataset. Privacy restrictions may re…
Optimized Data Pre-Processing for Discrimination Prevention
Flavio P. Calmon, Dennis Wei, Karthikeyan Natesan Ramamurthy +1
Non-discrimination is a recognized objective in algorithmic decision making. In this paper, we introduce a novel probabilistic formulation of data pre-processing for reducing discr…
Principal Inertia Components and Applications
Flavio P. Calmon, Ali Makhdoumi, Muriel Médard +3
We explore properties and applications of the Principal Inertia Components (PICs) between two discrete random variables and . The PICs lie in the intersection of information…
Strong Data Processing Inequalities for Input Constrained Additive Noise Channels
Flavio P. Calmon, Yury Polyanskiy, Yihong Wu
This paper quantifies the intuitive observation that adding noise reduces available information by means of non-linear strong data processing inequalities. Consider the random vari…
Multi-user guesswork and brute force security
Mark M. Christiansen, Ken R. Duffy, Flavio du Pin Calmon +1
The Guesswork problem was originally motivated by a desire to quantify computational security for single user systems. Leveraging recent results from its analysis, we extend the re…
Guessing a password over a wireless channel (on the effect of noise non-uniformity)
Mark M. Christiansen, Ken R. Duffy, Flavio du Pin Calmon +1
A string is sent over a noisy channel that erases some of its characters. Knowing the statistical properties of the string's source and which characters were erased, a listener tha…
Multi-Path TCP with Network Coding for Mobile Devices in Heterogeneous Networks
Jason Cloud, Flavio du Pin Calmon, Weifei Zeng +3
Existing mobile devices have the capability to use multiple network technologies simultaneously to help increase performance; but they rarely, if at all, effectively use these tech…
Brute force searching, the typical set and Guesswork
Mark M. Christiansen, Ken R. Duffy, Flavio du Pin Calmon +1
Consider the situation where a word is chosen probabilistically from a finite list. If an attacker knows the list and can inquire about each word in turn, then selecting the word v…