GMM
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Generative Modeling | GMM (d=40) | MaxSWD0.1 | 18 | |
| Toy target distribution sampling | GMM40 d = 50 | W2 (Entropy Regulated, eps=0.05)27.01 | 18 | |
| Sampling from synthetic distributions | 25GMM (D=2) | Delta Log Partition Function Error (Zr)0.003 | 13 | |
| Sampling | GMM-10 N=30 chains | R Value6,231 | 9 | |
| Sampling | GMM-10 (N=10 chains) | R681 | 9 | |
| Sampling | GMM-10 N=6 chains | R1,743 | 9 | |
| Generative Modeling | GMM d=100 | MaxSWD0.15 | 9 | |
| Jacobian Estimation for Boltzmann Sampling | 10D GMM | Mean Error0.0189 | 9 | |
| Amortised Sampling | GMM40 d=5 | Sinkhorn Distance83.3 | 7 | |
| Amortised Sampling | GMM40 d=2 | Sinkhorn Distance6.46 | 7 | |
| Synthetic Distribution Sampling | GMM20 | KSD3.012 | 7 | |
| Synthetic Distribution Sampling | GMM10 | KSD1.97 | 7 | |
| Synthetic Distribution Sampling | GMM5 | KSD1.375 | 7 | |
| Synthetic Distribution Sampling | GMM2 | KSD0.93 | 7 | |
| Continuous density estimation | GMM 30-D | FD1.05 | 7 | |
| Continuous density estimation | GMM 10-D | Fréchet Distance0.39 | 7 | |
| Toy target distribution sampling | GMM40 d = 2 | Entropy-Regularised W2 (ϵ=0.05)1.24 | 7 | |
| Boltzmann sampling | GMM-40 d = 2 | NLL6.96 | 5 | |
| Variational Inference | GMM 3 2 | RKL1.8 | 4 | |
| Dataset Summarization | GMM-100 | MMDw0.0282 | 4 | |
| Clustering | GMM K=7 (Mode B) | F1 Score0.86 | 4 | |
| Clustering | GMM K=5 (Mode B) | F1 Score0.93 | 4 | |
| Clustering | GMM (K=2) (Mode B) | F1 Score99 | 4 | |
| Clustering | GMM K=7 | F1 Score78 | 4 | |
| Clustering | GMM K=5 | F1 Score0.9 | 4 |
Showing 25 of 31 rows