This paper introduces a new theoretical framework for understanding signal recovery under sparse adversarial corruption in linear measurements. It moves beyond traditional exact recovery guarantees to characterize the information that remains robust even when exact recovery fails. The research proposes a method to recover this robust information set, which is defined by the kernel of a specific projection matrix related to the measurement matrix and the corruption sparsity level. AI
Summary written by gemini-2.5-flash-lite from 1 source. How we write summaries →
RANK_REASON This is a theoretical computer science paper published on arXiv. [lever_c_demoted from research: ic=1 ai=0.4]