001046979 001__ 1046979
001046979 005__ 20251007202036.0
001046979 0247_ $$2doi$$a10.48550/ARXIV.2507.10360
001046979 037__ $$aFZJ-2025-04047
001046979 041__ $$aEnglish
001046979 1001_ $$0P:(DE-HGF)0$$aFoos, Niklas$$b0
001046979 245__ $$aBeyond-mean-field fluctuations for the solution of constraint satisfaction problems
001046979 260__ $$barXiv$$c2025
001046979 3367_ $$0PUB:(DE-HGF)25$$2PUB:(DE-HGF)$$aPreprint$$bpreprint$$mpreprint$$s1759844041_9312
001046979 3367_ $$2ORCID$$aWORKING_PAPER
001046979 3367_ $$028$$2EndNote$$aElectronic Article
001046979 3367_ $$2DRIVER$$apreprint
001046979 3367_ $$2BibTeX$$aARTICLE
001046979 3367_ $$2DataCite$$aOutput Types/Working Paper
001046979 520__ $$aConstraint Satisfaction Problems (CSPs) lie at the heart of complexity theory and find application in a plethora of prominent tasks ranging from cryptography to genetics. Classical approaches use Hopfield networks to find approximate solutions while recently, modern machine-learning techniques like graph neural networks have become popular for this task. In this study, we employ the known mapping of MAX-2-SAT, a class of CSPs, to a spin-glass system from statistical physics, and use Glauber dynamics to approximately find its ground state, which corresponds to the optimal solution of the underlying problem. We show that Glauber dynamics outperforms the traditional Hopfield-network approach and can compete with state-of-the-art solvers. A systematic theoretical analysis uncovers the role of stochastic fluctuations in finding CSP solutions: even in the absense of thermal fluctuations at $T=0$ a significant portion of spins, which correspond to the CSP variables, attains an effective spin-dependent non-zero temperature. These spins form a subspace in which the stochastic Glauber dynamics continuously performs flips to eventually find better solutions. This is possible since the energy is degenerate, such that spin flips in this free-spin space do not require energy. Our theoretical analysis leads to new deterministic solvers that effectively account for such fluctuations, thereby reaching state-of-the-art performance.
001046979 536__ $$0G:(DE-HGF)POF4-5214$$a5214 - Quantum State Preparation and Control (POF4-521)$$cPOF4-521$$fPOF IV$$x0
001046979 536__ $$0G:(BMBF)13N15685$$aVerbundprojekt: German Quantum Computer based on Superconducting Qubits (GEQCOS) - Teilvorhaben: Charakterisierung, Kontrolle und Auslese (13N15685)$$c13N15685$$x1
001046979 588__ $$aDataset connected to DataCite
001046979 650_7 $$2Other$$aDisordered Systems and Neural Networks (cond-mat.dis-nn)
001046979 650_7 $$2Other$$aFOS: Physical sciences
001046979 7001_ $$0P:(DE-HGF)0$$aEpping, Bastian$$b1
001046979 7001_ $$0P:(DE-Juel1)198809$$aGrundler, Jannik$$b2$$ufzj
001046979 7001_ $$0P:(DE-Juel1)202049$$aCiobanu, Alexandru$$b3$$ufzj
001046979 7001_ $$0P:(DE-Juel1)200194$$aSingh, Ajainderpal$$b4
001046979 7001_ $$0P:(DE-Juel1)195623$$aBode, Tim$$b5$$ufzj
001046979 7001_ $$0P:(DE-Juel1)144806$$aHelias, Moritz$$b6$$ufzj
001046979 7001_ $$0P:(DE-Juel1)156459$$aDahmen, David$$b7$$ufzj
001046979 773__ $$a10.48550/ARXIV.2507.10360
001046979 909CO $$ooai:juser.fz-juelich.de:1046979$$pVDB
001046979 9101_ $$0I:(DE-588b)5008462-8$$6P:(DE-Juel1)198809$$aForschungszentrum Jülich$$b2$$kFZJ
001046979 9101_ $$0I:(DE-588b)5008462-8$$6P:(DE-Juel1)202049$$aForschungszentrum Jülich$$b3$$kFZJ
001046979 9101_ $$0I:(DE-588b)5008462-8$$6P:(DE-Juel1)195623$$aForschungszentrum Jülich$$b5$$kFZJ
001046979 9101_ $$0I:(DE-588b)5008462-8$$6P:(DE-Juel1)144806$$aForschungszentrum Jülich$$b6$$kFZJ
001046979 9101_ $$0I:(DE-588b)5008462-8$$6P:(DE-Juel1)156459$$aForschungszentrum Jülich$$b7$$kFZJ
001046979 9131_ $$0G:(DE-HGF)POF4-521$$1G:(DE-HGF)POF4-520$$2G:(DE-HGF)POF4-500$$3G:(DE-HGF)POF4$$4G:(DE-HGF)POF$$9G:(DE-HGF)POF4-5214$$aDE-HGF$$bKey Technologies$$lNatural, Artificial and Cognitive Information Processing$$vQuantum Materials$$x0
001046979 9141_ $$y2025
001046979 920__ $$lyes
001046979 9201_ $$0I:(DE-Juel1)PGI-12-20200716$$kPGI-12$$lQuantum Computing Analytics$$x0
001046979 980__ $$apreprint
001046979 980__ $$aVDB
001046979 980__ $$aI:(DE-Juel1)PGI-12-20200716
001046979 980__ $$aUNRESTRICTED