2 papers
stat.ML2026
Tail Annealing for Heavy-Tailed Flow Matching
Jean Pachebat
Standard generative models struggle with heavy-tailed data: Lipschitz architectures cannot produce power-law tails from Gaussian noise, and interpolating between heavy-tailed data…
stat.ML2025
Iterative Tilting for Diffusion Fine-Tuning
Jean Pachebat, Giovanni Conforti, Alain Durmus +1
We introduce iterative tilting, a gradient-free method for fine-tuning diffusion models toward reward-tilted distributions. The method decomposes a large reward tilt in…