CoDe: Blockwise Control for Denoising Diffusion Models
This work addresses the challenge of efficient inference-time control for diffusion models, offering a gradient-free alternative that is incremental but practical for users needing task-specific alignment.
The paper tackles the problem of aligning diffusion models to downstream tasks without requiring finetuning or differentiable guidance, by proposing CoDe, a blockwise sampling method that achieves competitive performance in reward alignment and prompt instruction following.
Aligning diffusion models to downstream tasks often requires finetuning new models or gradient-based guidance at inference time to enable sampling from the reward-tilted posterior. In this work, we explore a simple inference-time gradient-free guidance approach, called controlled denoising (CoDe), that circumvents the need for differentiable guidance functions and model finetuning. CoDe is a blockwise sampling method applied during intermediate denoising steps, allowing for alignment with downstream rewards. Our experiments demonstrate that, despite its simplicity, CoDe offers a favorable trade-off between reward alignment, prompt instruction following, and inference cost, achieving a competitive performance against the state-of-the-art baselines. Our code is available at: https://github.com/anujinho/code.