REVIEW 9 cited by
Spatial Reasoning with Denoising Models
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
Spatial Reasoning with Denoising Models
read the original abstract
We introduce Spatial Reasoning Models (SRMs), a framework to perform reasoning over sets of continuous variables via denoising generative models. SRMs infer continuous representations on a set of unobserved variables, given observations on observed variables. Current generative models on spatial domains, such as diffusion and flow matching models, often collapse to hallucination in case of complex distributions. To measure this, we introduce a set of benchmark tasks that test the quality of complex reasoning in generative models and can quantify hallucination. The SRM framework allows to report key findings about importance of sequentialization in generation, the associated order, as well as the sampling strategies during training. It demonstrates, for the first time, that order of generation can successfully be predicted by the denoising network itself. Using these findings, we can increase the accuracy of specific reasoning tasks from <1% to >50%. Our project website provides additional videos, code, and the benchmark datasets: https://geometric-rl.mpi-inf.mpg.de/srm
Forward citations
Cited by 9 Pith papers
-
When Do Diffusion Models learn to Generate Multiple Objects?
Diffusion models' multi-object generation is limited primarily by scene complexity and held-out combinations rather than imbalance, with counting difficult in low data and compositional generalization collapsing as mo...
-
AsyncPatch Diffusion: spatially-flexible image generation
AsyncPatch Diffusion introduces asynchronous per-region noise levels in diffusion models, proves a valid ELBO, and uses a controlled sampler to support spatially adaptive generation and native inpainting.
-
Inference-Time Scaling in Diffusion Models through Iterative Partial Refinement
IPR improves valid solution rates on MNIST Sudoku from 55.8% to 75.0% by iteratively refining partial regions in sequential diffusion models without external verifiers or reward models.
-
Visual Diffusion Models are Geometric Solvers
Standard visual diffusion models operating in pixel space can approximate solutions to the inscribed square, Steiner tree, and simple polygon problems.
-
From Interface to Inference: Eliciting Any-Order Inference from Any-Order Models
Masked diffusion language models fail to exploit their any-order interface because of positional uncertainty; insertion-based (FlexMDM) and latent-segment (LatentMDM) variants recover distinct any-order inference beha...
-
DiBS: Diffusion-Informed Branch Selection
DiBS ranks candidate values with a diffusion model inside a symbolic Sudoku solver to reduce nodes, backtracks, and long-tail search cost on the Royle 17-clue benchmark.
-
Score-Control for Hallucination Reduction in Diffusion Models
VSM modulates the score Jacobian using variance guidance to reduce hallucinations in diffusion models by up to 25% on synthetic and real datasets while preserving fidelity and diversity.
-
When Do Diffusion Models learn to Generate Multiple Objects?
Using the mosaic controlled dataset framework, experiments show scene complexity dominates over concept imbalance in diffusion model failures for multi-object generation, with counting especially hard in low-data regi...
-
Denoising, Fast and Slow: Difficulty-Aware Adaptive Sampling for Image Generation
Patch Forcing enables diffusion models to denoise image patches at varying rates based on predicted difficulty, advancing easier regions first to improve context and achieve better generation quality on ImageNet while...
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.