Under the direction of National cooperative for the Disposal of radioactive Waste (NAGrA), a probabilistic seismic hazard analysis was conducted for the swiss nuclear power plant sites. the study has become known under the name "PEGAsOs Project." this is the first of a group of papers in this volume that describes the seismic source characterization methodology and the main results of the project. A formal expert elicitation process was used, including dissemination of a comprehensive database, multiple workshops for identification and discussion of alternative models and interpretations, elicitation interviews, feedback to provide the experts with the implications of their preliminary assessments, and full documentation of the assessments. A number of innovative approaches to the seismic source characterization methodology were developed by four expert groups and implemented in the study. the identification of epistemic uncertainties and treatment using logic trees were important elements of the assessments. relative to the assessment of the seismotectonic framework, the four expert teams identified similar main seismotectonic elements: the rhine Graben, the Jura / Molasse regions, Helvetic and crystalline subdivisions of the Alps, and the southern Germany region. In
Existing neural network verifiers compute a proof that each input is handled correctly under a given perturbation by propagating a symbolic abstraction of reachable values at each layer. This process is repeated from scratch independently for each input (e.g., image) and perturbation (e.g., rotation), leading to an expensive overall proof effort when handling an entire dataset. In this work, we introduce a new method for reducing this verification cost without losing precision based on a key insight that abstractions obtained at intermediate layers for different inputs and perturbations can overlap or contain each other. Leveraging our insight, we introduce the general concept of shared certificates, enabling proof effort reuse across multiple inputs to reduce overall verification costs. We perform an extensive experimental evaluation to demonstrate the effectiveness of shared certificates in reducing the verification cost on a range of datasets and attack specifications on image classifiers including the popular patch and geometric perturbations. We release our implementation at https://github.com/eth-sri/proof-sharing.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.