Research map / Image generation / Autoregressive and latent generation
Dream the Impossible: Outlier Imagination with Diffusion Models
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| replace Gaussian noise on token embeddings with low-likelihood latent sampling (ablation) | Dream-OOD · ImageNet-100 avg | FPR95: 41.35 -> 38.76 | Helped |
| replace learnable noise with latent sampling (ablation) | Dream-OOD · CIFAR-100 avg | FPR95: 48.05 -> 40.31 | Helped |
| replace embedding interpolation with latent sampling (ablation) | Dream-OOD · CIFAR-100 avg | AUROC: 87.09 -> 90.15 | Helped |
| replace disjoint class-name prompts with latent sampling (ablation) | Dream-OOD · ImageNet-100 avg | FPR95: 43.55 -> 38.76 | Helped |
| raise OOD regularization weight beta beyond 1.0 (ablation) | Dream-OOD · ImageNet-100 | OOD detection | Hurt |
| increase Gaussian kernel variance to 0.2 (ablation) | Dream-OOD · ImageNet-100 | OOD detection | Hurt |
| vary k in k-NN distance from 100 to 500 (ablation) | Dream-OOD · ImageNet-100 | OOD detection | No clear effect |
| use latent-sampled inliers instead of generic prompts for synthetic ID data (ablation) | Dream-ID · ImageNet-100 test | accuracy: 87.74 -> 88.46 | Helped |
| use latent-sampled inliers instead of generic prompts (ablation) | Dream-ID · ImageNet-A | accuracy: 11.18 -> 12.13 | Helped |
Closest papers by meaning
- Diffusion-based Semantic Outlier Generation via Nuisance Awareness for Out-of-Distribution Detection (AAAI 2025 · similarity 0.91)
- Out-of-Distribution Detection with Prototypical Outlier Proxy (AAAI 2025 · similarity 0.88)
- Image Background Serves as Good Proxy for Out-of-distribution Data (ICLR 2024 · similarity 0.88)
- ALTBI: Constructing Improved Outlier Detection Models via Optimization of Inlier-Memorization Effect (AAAI 2025 · similarity 0.88)
- Envisioning Outlier Exposure by Large Language Models for Out-of-Distribution Detection (ICML 2024 · similarity 0.88)
- BOOD: Boundary-based Out-Of-Distribution Data Generation (ICML 2025 · similarity 0.88)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.