AIM
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Image Matting | AIM-500 | SAD9.089 | 24 | |
| Super-Resolution | AIM (val) | PSNR22.7 | 17 | |
| Affective Image Manipulation | AIM | CLIP-T0.2216 | 14 | |
| Super-Resolution | AIM Track 2 2019 | PSNR23.37 | 14 | |
| Jailbreak Defense | AIM AdvE | Attack Success Rate (ASR)0 | 14 | |
| Adaptive Influence Maximization | AIM (N=500 graphs, T=20, K=10) (test) | Average Reward161.71 | 9 | |
| Adaptive Influence Maximization | AIM (N=500 graphs, T=10, K=70) (test) | Average Reward324.15 | 9 | |
| Adaptive Influence Maximization | AIM (N=500 graphs, T=10, K=60) (test) | Average Reward303.5 | 9 | |
| Adaptive Influence Maximization | AIM (N=500 graphs, T=10, K=50) (test) | Average Reward286.11 | 9 | |
| Influence Maximization | AIM N=2500 (test) | Avg. Reward654.29 | 9 | |
| Influence Maximization | AIM N=2000 (test) | Avg. Reward605.55 | 9 | |
| Influence Maximization | AIM N=1500 (test) | Average Reward540.78 | 9 | |
| Influence Maximization | AIM N=1000 (test) | Average Reward416.06 | 9 | |
| SAT Solving | AIM | MRPP r˜1.2 | 9 | |
| Novel View and Pose Synthesis | AiM Zebra (test) | PSNR15.54 | 6 | |
| Novel View and Pose Synthesis | AiM Horse subset (test) | PSNR16.19 | 6 | |
| 4D equine reconstruction | AiM | PCK@0.0584.2 | 6 | |
| SAT Solving | AIM structured SAT | MRPP r-tilde1.03 | 6 | |
| Planning | AIM N=500, K=70 | Performance324.15 | 5 | |
| Text-to-image | AIM-500 | CLIP Score85.2 | 5 | |
| Average Reward | AIM (N=200, T=20, K=10) | Average Reward86.16 | 2 | |
| Average Reward | AIM (N=200, T=10, K=30) | Average Reward137.28 | 2 | |
| Average Reward | AIM (N=200, T=10, K=20) | Average Reward123.83 | 2 | |
| Average Reward | AIM N=200, T=10, K=10 | Average Reward83.21 | 2 |
Showing 24 of 24 rows