Research map / Image generation / Fast diffusion sampling
Hot Fuzz: Temperature-Tunable Composition of Diffusion models with Fuzzy Logic
Peter Blohm, Vikas K Garg
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 |
|---|---|---|---|
| use Dombi composition instead of PoE/MoE | Diffusion composition · SAT experiment XOR2 | Sat: 0.00 -> 0.97 | Improved |
| use Dombi composition instead of PoE/MoE | Diffusion composition · SAT experiment Maj2 | Unif: 0.80 -> 1.00 | Improved |
| use Dombi composition on formula with many negations (k=10) | Diffusion composition · SAT experiment OneHot10 | Sat: 0.07 | Hurt |
| use Dombi conjunction instead of SuperDiff and | Stable Diffusion · Conjunctive prompts | CLIP: 24.87 -> 25.50 | Improved |
| use Dombi conjunction instead of scaled PoE | Stable Diffusion · Conjunctive prompts | ImageReward: -1.55 -> -1.18 | Improved |
| use Dombi instead of ICN on contrastive prompts | Stable Diffusion · Contrastive prompts | CLIP difference: 7.29 -> 7.40 | No clear effect |
| add FKC correction (gamma=2) (ablation) | Dombi composition · Multi-target docking | P1*P2 docking score: 71.36 -> 81.63 | Improved |
| add FKC correction (gamma=1) (ablation) | Dombi composition · Multi-target docking | P1*P2 docking score: 68.60 -> 72.83 | Improved |
| add FKC correction (ablation) | Dombi composition · Multi-target docking | diversity: 0.88 -> 0.86 | Mixed |
Closest papers by meaning
- Derivative-Free Guidance in Continuous and Discrete Diffusion Models with Soft Value-based Decoding (NeurIPS 2025 · similarity 0.86)
- The Superposition of Diffusion Models Using the Itô Density Estimator (ICLR 2025 · similarity 0.86)
- How Diffusion Models Learn to Factorize and Compose (NeurIPS 2024 · similarity 0.86)
- Scalable Discrete Diffusion Samplers: Combinatorial Optimization and Statistical Physics (ICLR 2025 · similarity 0.86)
- CoInD: Enabling Logical Compositions in Diffusion Models (ICLR 2025 · similarity 0.86)
- Fast Solvers for Discrete Diffusion Models: Theory and Applications of High-Order Algorithms (NeurIPS 2025 · similarity 0.85)
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.
