Outcome Indistinguishability
Outcome indistinguishability focuses on determining whether the outputs of two systems or algorithms are statistically distinguishable, a crucial concept in areas like cryptography, machine learning, and privacy-preserving data analysis. Current research investigates this through various lenses, including deep learning models (e.g., residual networks) for analyzing ciphertexts and embedding spaces for evaluating the impact of backdoor attacks on pre-trained models. The ability to assess outcome indistinguishability has significant implications for evaluating the security of cryptographic systems, the robustness of machine learning models, and the development of privacy-preserving techniques for data analysis.
Papers
October 11, 2024
May 30, 2024
January 29, 2024
November 19, 2023
September 3, 2023
May 23, 2023
February 24, 2023
January 21, 2023
November 25, 2022
October 16, 2022
March 9, 2022