Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Noise Hypernetworks: Amortizing Test-Time Compute in Diffusion Models

About

The new paradigm of test-time scaling has yielded remarkable breakthroughs in Large Language Models (LLMs) (e.g. reasoning models) and in generative vision models, allowing models to allocate additional computation during inference to effectively tackle increasingly complex problems. Despite the improvements of this approach, an important limitation emerges: the substantial increase in computation time makes the process slow and impractical for many applications. Given the success of this paradigm and its growing usage, we seek to preserve its benefits while eschewing the inference overhead. In this work we propose one solution to the critical problem of integrating test-time scaling knowledge into a model during post-training. Specifically, we replace reward guided test-time noise optimization in diffusion models with a Noise Hypernetwork that modulates initial input noise. We propose a theoretically grounded framework for learning this reward-tilted distribution for distilled generators, through a tractable noise-space objective that maintains fidelity to the base model while optimizing for desired characteristics. We show that our approach recovers a substantial portion of the quality gains from explicit test-time optimization at a fraction of the computational cost. Code is available at https://github.com/ExplainableML/HyperNoise

Luca Eyring, Shyamgopal Karthik, Alexey Dosovitskiy, Nataniel Ruiz, Zeynep Akata• 2025

Related benchmarks

TaskDatasetResultRank
Text-to-Image GenerationDPG-Bench
DPG Score72.4
156
Video GenerationVBench--
126
Super-ResolutionFFHQ 256 x 256
PSNR27.86
52
Super-ResolutionImageNet 256
PSNR21.6
50
InpaintingImageNet 256
PSNR19.55
30
Text-to-Image GenerationDrawBench
HPSv231.27
27
Text to ImagePick-a-Pic
HPSv232.31
15
HDR ReconstructionFFHQ 256 x 256
PSNR22.92
14
HDR ReconstructionImageNet 256 x 256
PSNR19.33
13
Text-to-Image GenerationGenEval 2
Object Accuracy83.59
12
Showing 10 of 25 rows

Other info

Follow for update