[2509.21513] DistillKac: Few-Step Image Generation via Damped Wave Equations

Computer Science > Machine Learning arXiv:2509.21513 (cs) [Submitted on 25 Sep 2025 (v1), last revised 2 Mar 2026 (this version, v3)] Title:DistillKac: Few-Step Image Generation via Damped Wave Equations Authors:Weiqiao Han, Chenlin Meng, Christopher D. Manning, Stefano Ermon View a PDF of the paper titled DistillKac: Few-Step Image Generation via Damped Wave Equations, by Weiqiao Han and 3 other authors View PDF HTML (experimental) Abstract:We present DistillKac, a fast image generator that uses the damped wave equation and its stochastic Kac representation to move probability mass at finite speed. In contrast to diffusion models whose reverse time velocities can become stiff and implicitly allow unbounded propagation speed, Kac dynamics enforce finite speed transport and yield globally bounded kinetic energy. Building on this structure, we introduce classifier-free guidance in velocity space that preserves square integrability under mild conditions. We then propose endpoint only distillation that trains a student to match a frozen teacher over long intervals. We prove a stability result that promotes supervision at the endpoints to closeness along the entire path. Experiments demonstrate DistillKac delivers high quality samples with very few function evaluations while retaining the numerical stability benefits of finite speed probability flows. Comments: Subjects: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV); ...