Ewsey R. Obzherin, Valentin Khrulkov · 2026 · arXiv (Cornell University)
DOI: 10.48550/arxiv.2610.29475 · cited by 0 works (OpenAlex) · open-access PDF
A amostragem de ODE gerativa de baixo nível NFE é restrita não apenas pela ordem do solvente, mas também pela forma como um orçamento fixo de avaliações de rede é distribuído ao longo do fluxo. Desenvolvemos um quadro de erro retroativo de nível de densidade que identifica o componente do defeito de equação modificada de um solvente que altera a margem evolutiva, excluindo a distorção da trajetória solenoide que é invisível na geometria tangente fixa de Wasserstein. O residual resultante induz uma regra de alocação de computação consciente da ordem do solvente e, através da liberdade de medição, um objetivo para selecionar representantes numericamente favoráveis da velocidade do mesmo caminho de densidade. Isso separa nossa abordagem do esquema de busca impulsionado por
Machine translation (Meta M2M100); the English original follows.
Low-NFE generative ODE sampling is constrained not only by solver order, but also by how a fixed budget of network evaluations is distributed along the flow. We develop a density-level backward-error framework that identifies the component of a solver's modified-equation defect that changes the evolving marginal, while excluding solenoidal trajectory distortion that is invisible in Wasserstein tangent geometry. The resulting residual induces a solver-order-aware compute-allocation rule and, through gauge freedom, an objective for selecting numerically favorable velocity representatives of the same density path. This separates our approach from schedule search driven by trajectory or endpoint surrogates and from higher-order solver design. Exact Gaussian and mixture models establish the mechanism, including cases where curvature-based allocation fails and continuum-optimal transport is not discretization-optimal. On a pretrained CIFAR-10 EDM probability-flow ODE, density-aware allocation consistently improves fixed-update Euler sampling in the low-NFE regime. Matched-seed convergence to a dense Euler reference confirms that the method accelerates approximation of the same high-NFE flow rather than changing the asymptotic sampler.
Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese
Topics: Trajectory, Solver, Residual, Mathematics, Convergence (economics), Applied mathematics
GET https://store.agentexchange.work/gw/papers/verify?title=...&year=... (0.003 USD), fetch this record as JSON with /gw/papers/paper?id=2610.29475, or list what it cites with /gw/papers/references?id=.... Pay per call in USDC (x402) or with a 19 USD card key for 1,000 calls. Catalogue · llms.txtSources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange