Research map / Image generation / Fast diffusion sampling
Denoising Likelihood Score Matching for Conditional Score-based Data Generation
Chen-Hao Chao, Wei-Fang Sun, Bo-Wun Cheng, Yi-Chen Lo, Chia-Che Chang, Yu-Lun Liu, Yu-Lin Chang, Chia-Ping Chen, Chun-Yi Lee
This paper tried a common change: Add a reconstruction / auxiliary loss
Add cross-entropy as an auxiliary loss to DLSM' · Inter-twining moon · training stability Improved
What 987 accepted papers report for this change
- 882 improved
- 41 no clear effect
- 20 mixed
- 44 hurt
- OPTION: Optimal Transport–Guided Flow Matching for Incomplete and Unaligned Multi-View Clustering ICML 2026 · add reconstruction loss L_rec · Scene-15 (aligned, 70% missing) · ACC: 32.4 -> 37.5Improved
- Detached Skip-Links and $R$-Probe: Decoupling Feature Aggregation from Gradient Propagation for MLLM OCR ICML 2026 · add R-probe auxiliary loss · OCRBench · score: 714.0 -> 721.0Improved
- Rethinking Time-Series Imputation as Conditional Inference along Temporal Evolution ICML 2026 · add reconstruction loss on observed entries (joint vs missing-only supervision) · ETTh1 10% missing · Masked MAE: 0.2002 -> 0.1658Improved
See this change in detailMatched by the wording of each paper's own result, read by AI from the full text. Papers report gains more often than failures, so this shows what gets published, not the chance a change works.
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 |
|---|---|---|---|
| Train the classifier with DLSM and cross-entropy instead of cross-entropy alone | Conditional score-based generation · Cifar-10 | FID: 4.10 -> 2.25 | Improved |
| Train the classifier with DLSM and cross-entropy instead of cross-entropy alone | Conditional score-based generation · Cifar-100 | FID: 4.52 -> 3.86 | Improved |
| Train the classifier with DLSM and cross-entropy instead of cross-entropy alone | Conditional score-based generation · Cifar-10 | IS: 9.08 -> 9.90 | Improved |
| Train the classifier with DLSM and cross-entropy instead of cross-entropy alone | Conditional score-based generation · Cifar-100 | IS: 11.53 -> 11.62 | Improved |
| Scale the likelihood score | Base conditional score-based method · Cifar-10 | FID: 4.10 -> 12.48 | Hurt |
| Scale the likelihood score | Base conditional score-based method · Cifar-100 | Recall: 0.62 -> 0.52 | Hurt |
| Add DLSM to cross-entropy classifier training | Classifier in conditional score-based generation · Inter-twining moon | likelihood score error: 0.190 -> 0.052 (c1); 0.181 -> 0.052 (c2) | Improved |
| Add cross-entropy as an auxiliary loss to DLSM' (ablation) | DLSM'-trained classifier · Inter-twining moon | training stability | Improved |
| Use cross-entropy alone instead of adding the DLSM' score-matching objective (ablation) | Classifier in conditional score-based generation · Inter-twining moon | likelihood score error | Hurt |
| Add the DLSM' score-matching objective to cross-entropy training (ablation) | Classifier in conditional score-based generation · Inter-twining moon | likelihood score error | Improved |
| Use DLSM' alone without the cross-entropy auxiliary loss (ablation) | Classifier in conditional score-based generation · Inter-twining moon | training stability | Hurt |
Closest papers by meaning
- Label-Noise Robust Diffusion Models (ICLR 2024 · similarity 0.86)
- What’s the score? Automated Denoising Score Matching for Nonlinear Diffusions (ICML 2024 · similarity 0.86)
- Theory on Score-Mismatched Diffusion Models and Zero-Shot Conditional Samplers (ICLR 2025 · similarity 0.86)
- Expressive Score-Based Priors for Distribution Matching with Geometry-Preserving Regularization (ICML 2025 · similarity 0.86)
- Sampling Binary Data by Denoising through Score Functions (ICML 2025 · similarity 0.86)
- Training Diffusion-based Generative Models with Limited Data (ICML 2025 · similarity 0.86)
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.
