Matches in SemOpenAlex for { <https://semopenalex.org/work/W4376167265> ?p ?o ?g. }
Showing items 1 to 74 of
74
with 100 items per page.
- W4376167265 abstract "Deep neural networks (DNNs) are increasingly being deployed to perform safety-critical tasks. The opacity of DNNs, which prevents humans from reasoning about them, presents new safety and security challenges. To address these challenges, the verification community has begun developing techniques for rigorously analyzing DNNs, with numerous verification algorithms proposed in recent years. While a significant amount of work has gone into developing these verification algorithms, little work has been devoted to rigorously studying the computability and complexity of the underlying theoretical problems. Here, we seek to contribute to the bridging of this gap. We focus on two kinds of DNNs: those that employ piecewise-linear activation functions (e.g., ReLU), and those that employ piecewise-smooth activation functions (e.g., Sigmoids). We prove the two following theorems: 1) The decidability of verifying DNNs with a particular set of piecewise-smooth activation functions is equivalent to a well-known, open problem formulated by Tarski; and 2) The DNN verification problem for any quantifier-free linear arithmetic specification can be reduced to the DNN reachability problem, whose approximation is NP-complete. These results answer two fundamental questions about the computability and complexity of DNN verification, and the ways it is affected by the network's activation functions and error tolerance; and could help guide future efforts in developing DNN verification tools." @default.
- W4376167265 created "2023-05-12" @default.
- W4376167265 creator A5026961968 @default.
- W4376167265 creator A5044509944 @default.
- W4376167265 creator A5052302712 @default.
- W4376167265 creator A5081405133 @default.
- W4376167265 date "2023-05-10" @default.
- W4376167265 modified "2023-09-23" @default.
- W4376167265 title "DNN Verification, Reachability, and the Exponential Function Problem" @default.
- W4376167265 doi "https://doi.org/10.48550/arxiv.2305.06064" @default.
- W4376167265 hasPublicationYear "2023" @default.
- W4376167265 type Work @default.
- W4376167265 citedByCount "0" @default.
- W4376167265 crossrefType "posted-content" @default.
- W4376167265 hasAuthorship W4376167265A5026961968 @default.
- W4376167265 hasAuthorship W4376167265A5044509944 @default.
- W4376167265 hasAuthorship W4376167265A5052302712 @default.
- W4376167265 hasAuthorship W4376167265A5081405133 @default.
- W4376167265 hasBestOaLocation W43761672651 @default.
- W4376167265 hasConcept C110251889 @default.
- W4376167265 hasConcept C111498074 @default.
- W4376167265 hasConcept C11413529 @default.
- W4376167265 hasConcept C134306372 @default.
- W4376167265 hasConcept C136643341 @default.
- W4376167265 hasConcept C151376022 @default.
- W4376167265 hasConcept C152062344 @default.
- W4376167265 hasConcept C153269930 @default.
- W4376167265 hasConcept C164155591 @default.
- W4376167265 hasConcept C168773769 @default.
- W4376167265 hasConcept C17095337 @default.
- W4376167265 hasConcept C199360897 @default.
- W4376167265 hasConcept C2524010 @default.
- W4376167265 hasConcept C2777669093 @default.
- W4376167265 hasConcept C32278780 @default.
- W4376167265 hasConcept C33923547 @default.
- W4376167265 hasConcept C41008148 @default.
- W4376167265 hasConcept C49016629 @default.
- W4376167265 hasConcept C80444323 @default.
- W4376167265 hasConceptScore W4376167265C110251889 @default.
- W4376167265 hasConceptScore W4376167265C111498074 @default.
- W4376167265 hasConceptScore W4376167265C11413529 @default.
- W4376167265 hasConceptScore W4376167265C134306372 @default.
- W4376167265 hasConceptScore W4376167265C136643341 @default.
- W4376167265 hasConceptScore W4376167265C151376022 @default.
- W4376167265 hasConceptScore W4376167265C152062344 @default.
- W4376167265 hasConceptScore W4376167265C153269930 @default.
- W4376167265 hasConceptScore W4376167265C164155591 @default.
- W4376167265 hasConceptScore W4376167265C168773769 @default.
- W4376167265 hasConceptScore W4376167265C17095337 @default.
- W4376167265 hasConceptScore W4376167265C199360897 @default.
- W4376167265 hasConceptScore W4376167265C2524010 @default.
- W4376167265 hasConceptScore W4376167265C2777669093 @default.
- W4376167265 hasConceptScore W4376167265C32278780 @default.
- W4376167265 hasConceptScore W4376167265C33923547 @default.
- W4376167265 hasConceptScore W4376167265C41008148 @default.
- W4376167265 hasConceptScore W4376167265C49016629 @default.
- W4376167265 hasConceptScore W4376167265C80444323 @default.
- W4376167265 hasLocation W43761672651 @default.
- W4376167265 hasLocation W43761672652 @default.
- W4376167265 hasOpenAccess W4376167265 @default.
- W4376167265 hasPrimaryLocation W43761672651 @default.
- W4376167265 hasRelatedWork W148542191 @default.
- W4376167265 hasRelatedWork W1850275321 @default.
- W4376167265 hasRelatedWork W2126053898 @default.
- W4376167265 hasRelatedWork W2145835888 @default.
- W4376167265 hasRelatedWork W2397752352 @default.
- W4376167265 hasRelatedWork W2595598880 @default.
- W4376167265 hasRelatedWork W2810777846 @default.
- W4376167265 hasRelatedWork W3014139035 @default.
- W4376167265 hasRelatedWork W3023596811 @default.
- W4376167265 hasRelatedWork W4376167265 @default.
- W4376167265 isParatext "false" @default.
- W4376167265 isRetracted "false" @default.
- W4376167265 workType "article" @default.