Quentin Bertrand

Beyond the Manifold Hypothesis: Hybrid Spectral Parameterizations for Flow Matching

Flow matching and diffusion can be trained to predict different quantities, most commonly the data $x_1$, the source noise $x_0$, or the velocity~$v$. Although theoretically equivalent, these can lead to substantially different performances. We identify two main drivers for these differences: the source--data signal-to...