On Parallel Scalable Uniform SAT Witness Generation
PaperFirst seen 9/7/2026
Last seen 9/7/2026
Evidence 9 chunks
NEIGHBORHOOD
34 nodes · 48 edgesgraph · On Parallel Scalable Uniform SAT Witness Generation · depth=1
RELATIONSHIPS
35 connectionsThe paper evaluates UniGen2 extensively through experiments on diverse benchmarks.
The paper mentions Jerrum, Valiant, and Vazirani as pioneers of uniform SAT witness generation.
The paper mentions Valiant as a pioneer of uniform SAT witness generation.
The paper mentions Vazirani as a pioneer of uniform SAT witness generation.
The paper compares UniGen2 against UniGen in experiments.
The paper evaluates the parallel version of UniGen2 for speedup and uniformity.
The paper focuses on sampling satisfying assignments of CNF formulae.
The paper uses independent support to restrict the sampling set.
The paper addresses functional verification as its application domain.
The paper is focused on uniform stimulus generation as a key problem.
The paper addresses validation of designs under verification.
The paper is affiliated with Rice University.
The paper uses r-wise independent hash functions as a theoretical basis.
The paper introduces UniGen2 as a new algorithm for uniform SAT witness sampling.
The paper presents and evaluates parallel UniGen2 as a parallel implementation of UniGen2.
The paper mentions the end of Dennard scaling as motivation for parallelization.
The paper acknowledges Armando Solar-Lezama for providing benchmarks.
The paper acknowledges Mate Soos for tweaking CryptoMiniSAT to support UniGen2.
The paper addresses approximate model counting as a related application of SAT witness sampling.
The paper cites RACE as an interval propagation based stimulus generation tool.
The paper discusses interval propagation as an industrial technique for stimulus generation.
The paper discusses MCMC-based methods for generating samples.
The paper discusses WBDD-based sampling as an alternative approach.
The paper mentions conversion of constraints into belief networks as a related approach.
The paper mentions UniWit as a predecessor to UniGen.
The paper uses universal hashing as a key technique for partitioning witness space.
The paper is authored by Supratik Chakraborty.
The paper mentions random seeding of SAT solvers as a heuristic baseline approach.
The paper is authored by Daniel J. Fremont.
The paper is authored by Kuldeep S. Meel.
The paper is authored by Sanjit A. Seshia.
The paper is authored by Moshe Y. Vardi.
The paper uses ISCAS89 benchmark circuits as evaluation benchmarks.
The paper uses bit-blasted SMTLib benchmarks for evaluation.
The paper uses constraints from bounded model checking as evaluation benchmarks.