aiqbench analysis
This paper establishes a theoretical bridge between diffusion sampling and imaginary-time quantum annealing, framing score-based learning errors as Hamiltonian perturbations within a supersymmetric framework. By applying Morita–Nishimori boundary cancellation principles, the authors demonstrate a method to significantly suppress residual Hellinger error in reverse stochastic differential equation benchmarks. The work advances the mathematical understanding of how quantum-inspired dynamics can optimize sampling schedules for complex probability densities.
Boundary-Cancelled Score-Hamiltonian Sampling
arXiv:2609.05904v1 Announce Type: new Abstract: Diffusion sampling can be viewed as imaginary-time annealing of probability densities. From a forward/backward Euclidean Schr\"odinger pair, we show that fixing the noising drift forces the reverse score term, and that the same logarithmic force is the one-sided imaginary-time counterdiabatic connection of a supersymmetric Score Hamiltonian. The correspondence turns score-learning error into a Hamiltonian perturbation and yields a schedule principle: if the first $r$ terminal derivatives vanish, Morita--Nishimori boundary cancellation suppresses the residual Hellinger error from $T^{-2}$ to $T^{-2r-2}$ until score or sampling floors are reached. Reverse-SDE benchmarks, including two-dimensional learned-score tests, confirm the predicted improvement.
Abstract:Diffusion sampling can be viewed as imaginary-time annealing of probability densities. From a forward/backward Euclidean Schrödinger pair, we show that fixing the noising drift forces the reverse score term, and that the same logarithmic force is the one-sided imaginary-time counterdiabatic connection of a supersymmetric Score Hamiltonian. The correspondence turns score-learning error into a Hamiltonian perturbation and yields a schedule principle: if the first $r$ terminal derivatives vanish, Morita--Nishimori boundary cancellation suppresses the residual Hellinger error from $T^{-2}$ to $T^{-2r-2}$ until score or sampling floors are reached. Reverse-SDE benchmarks, including two-dimensional learned-score tests, confirm the predicted improvement.
Quantum Physics (quant-ph); Disordered Systems and Neural Networks (cond-mat.dis-nn)
arXiv:2609.05904 [quant-ph]
(or arXiv:2609.05904v1 [quant-ph] for this version)
https://doi.org/10.48550/arXiv.2609.05904 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 196 words
Related coverage
Debugging Logprob Mismatches in LLM Reinforcement Learning
AMD · Sep 23, 2026
Handling Imbalanced Classification: What Works Better Than SMOTE
Analytics Vidhya · Jul 12, 2026
SCOPE Before You Sample: Conformal Scores Guide Posterior Exploration for Fast and Calibrated Bayesian Predictive Inference
arXiv · Oct 14, 2026
P2TPO: Structure-Aligned Page-to-Token Credit Assignment for Full-Page Document Parsing
arXiv · Oct 16, 2026