arXiv:2610.01933v1 Announce Type: new
Abstract: Inference-time scaling adapts pretrained diffusion models to new sampling tasks without additional training. Existing methods rely primarily on Monte C...
By Zuokai Wen, Louis Grenioux, Weinan E, Jiequn Han
The paper introduces a method for specificity‑aware diffusion steering that suppresses undesired samples while preserving desired ones. By formulating the problem as a target‑design task, it derives a time‑dependent target distribution based on overlap between positive and negative reference distributions, and samples from it using a variance‑reduced Sequential Monte Carlo (SMC) sampler. Experiments on synthetic, class‑contrastive, text‑to‑image, and peptide‑MHC tasks demonstrate reduced mode shift, improved sampling stability, and better suppression of undesired regions compared to negative‑guidance baselines.
By Luran Wang, Linrui Ma, Hannes St\"ark, Regina Barzilay
arXiv:2607. 20913v1 Announce Type: new Abstract: Fine-tuning large diffusion models for new domains or styles involves a trade-off: improving target-specific generation often degrades the pretrained model's broad generative capability.
By Yi Xiong, Yuan-Yuan Cheng, Xiao-Ming Fu
arXiv:2601. 21026v2 Announce Type: replace-cross Abstract: Sampling configurations at thermodynamic equilibrium is a central challenge in statistical physics.
By Louis Grenioux, Maxence Noble
The paper introduces a new Markov chain Monte Carlo method that samples from multimodal distributions by interpolating along the diffusion path of a noising diffusion process, preserving mode weights and improving mixing. It proposes a Metropolis-adjusted diffusion path (MAD-Path) sampler that corrects for bias from approximate score estimates and discretization errors, ensuring the target distribution remains invariant. Experiments on Bayesian posteriors demonstrate that MAD-Path outperforms tempering-based MCMC and unadjusted diffusion samplers in global exploration and accurate mode-weight estimation.
By Han Chen, Sifan Liu, Jun Yang
The paper introduces GRAS, a method that improves training‑free reward alignment for discrete diffusion models by reducing variance in guided proposals and adapting the resampling temperature during search. It achieves this without adding denoiser cost, using Rao‑Blackwellized estimates for differentiable rewards and a leave‑one‑out baseline for non‑differentiable ones. Experiments on regulatory DNA and protein design show GRAS outperforms existing training‑free techniques and rivals reward‑fine‑tuned models.
By Kwanyoung Kim