arXiv:2602. 08646v3 Announce Type: replace Abstract: We propose a gradient preconditioning method that makes reward-guided generation with one-step generative models both efficient and reliable.
Summary generated by The Flow from the publisher's feed. The full article lives at arXiv Machine Learning.
arXiv:2606. 18066v1 Announce Type: new Abstract: We introduce the Noise-Tilted Reverse Kernel (NTRK), a reward-guided diffusion sampler that injects reward gradients through the noise term, leaving the pretrained reverse kernel unchanged and requiring only a single sample per step.
arXiv:2607. 00691v1 Announce Type: new Abstract: Black-box optimization is a fundamental science and engineering tool that makes it possible to optimize objectives without gradient information.
Diffusion models have strong generative capabilities. However, their maximum likelihood training objective only focuses on reconstructing the data distribution, making it difficult to align with specific preferences.
arXiv:2603. 14504v2 Announce Type: replace-cross Abstract: Optimizing the noise samples of diffusion and flow models is an increasingly popular approach to align these models to target rewards at inference time.
arXiv:2606. 09635v1 Announce Type: cross Abstract: Ensuring the reliability of Large Language Models (LLMs) under distribution drift requires inference-time adaptation.
arXiv:2606. 28301v1 Announce Type: new Abstract: Inference-time scaling is a promising paradigm to improve generative models, especially when outputs must satisfy structural constraints or optimize downstream rewards.