Quantum Property Prediction on QM9 (RMSE)
0.0035HOMO-LUMO Gap (Delta_epsilon)Uni-Mol2
Evaluation results
| Method | Links | ||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Uni-Mol2Category=3D GNN, requires 3D conformer generation=true2026.05 | 0.0035 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Uni-MolCategory=3D GNN, requires 3D conformer generation=true2026.05 | 0.0047 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| MolSight (SigLIP2, S6)Category=Image-only, Configuration=single-model results2026.05 | 0.007 | — | — | — | — | — | — | 1.99 | — | — | — | — | — | — | — | — | — | — | — | — | |
| MolSight-BestCategory=Image-only, Configuration=best curriculum-trained architecture per task2026.05 | 0.007 | — | — | — | — | — | — | 1.99 | — | — | — | — | — | — | — | — | — | — | — | — | |
| PNA2024.02 | 0.139 | 0.535 | 0.476 | 0.099 | 0.109 | 28.503 | 0.032 | 31.644 | 25.038 | 27.338 | 22.308 | 0.179 | 0.353 | 0.361 | 0.373 | 1.007 | 0.256 | 0.277 | — | — | |
| PNA2024.02 | 0.14 | 0.53 | 0.48 | 0.1 | 0.11 | — | 28.5 | 0.03 | 0.03 | 31.64 | 27.34 | 0.18 | 0.36 | 0.37 | 0.31 | 1.01 | 0.26 | 0.28 | — | 0.35 | |
| GPS2024.02 | 0.15 | 0.94 | 0.6 | 0.11 | 0.11 | — | 30.42 | 0.03 | 0.03 | 14.5 | 12.92 | 0.19 | 0.33 | 0.36 | 0.36 | 1.42 | 0.16 | 0.12 | — | 0.33 | |
| ESA2024.02 | 0.15 | 0.56 | 0.4 | 0.1 | 0.11 | — | 28.33 | 0.03 | 0.03 | 4.78 | 6.02 | 0.16 | 0.24 | 0.25 | 0.22 | 0.75 | 0.08 | 0.05 | — | 0.24 | |
| ESA2024.02 | 0.152 | 0.564 | 0.398 | 0.103 | 0.114 | 28.328 | 0.026 | 4.777 | 6.799 | 6.018 | 6.104 | 0.158 | 0.241 | 0.243 | 0.245 | 0.746 | 0.079 | 0.05 | — | — | |
| GPS2024.02 | 0.154 | 0.945 | 0.605 | 0.113 | 0.108 | 30.421 | 0.033 | 14.496 | 15.82 | 12.923 | 13.081 | 0.19 | 0.33 | 0.334 | 0.363 | 1.422 | 0.158 | 0.124 | — | — | |
| GAT2024.02 | 0.16 | 0.55 | 0.48 | 0.11 | 0.13 | — | 30.91 | 0.03 | 0.03 | 27.82 | 24.92 | 0.4 | 0.48 | 0.38 | 0.34 | 0.9 | 0.19 | 0.27 | — | 0.41 | |
| GATv22024.02 | 0.16 | 0.55 | 0.46 | 0.1 | 0.12 | — | 30.15 | 0.03 | 0.03 | 25.4 | 23.42 | 0.33 | 0.33 | 0.38 | 0.34 | 1.08 | 0.26 | 0.28 | — | 0.31 | |
| Graphormer2024.02 | 0.16 | 0.63 | 0.4 | 0.11 | 0.11 | — | 29.63 | 0.06 | 0.06 | 24.6 | 19.01 | 0.18 | 0.3 | 0.3 | 0.26 | 64.88 | 0.1 | 0.12 | — | 0.29 | |
| GAT2024.02 | 0.163 | 0.552 | 0.479 | 0.107 | 0.124 | 30.911 | 0.033 | 27.823 | 30.914 | 28.924 | 31.494 | 0.191 | 0.392 | 0.483 | 0.383 | 0.972 | 0.211 | 0.266 | — | — | |
| GATv22024.02 | 0.163 | 0.545 | 0.46 | 0.104 | 0.127 | 30.149 | 0.032 | 25.405 | 24.919 | 23.422 | 23.24 | 0.19 | 0.384 | 0.397 | 0.406 | 1.078 | 0.264 | 0.276 | — | — | |
| Graphormer2024.02 | 0.163 | 0.627 | 0.404 | 0.107 | 0.112 | 29.628 | 0.059 | 24.6 | 15.546 | 19.006 | 31.198 | 0.177 | 0.289 | 0.302 | 0.301 | 0.261 | 64.877 | 0.102 | — | — | |
| GIN2024.02 | 0.164 | 0.55 | 0.438 | 0.104 | 0.12 | 29.925 | 0.032 | 42.024 | 25.098 | 24.746 | 26.942 | 0.187 | 0.391 | 0.377 | 0.37 | 1.317 | 0.196 | 0.167 | — | — | |
| DropGIN2024.02 | 0.166 | 0.552 | 0.445 | 0.106 | 0.123 | 29.87 | 0.034 | 29.817 | 24.741 | 34.019 | 25.412 | 0.195 | 0.378 | 0.404 | 0.376 | 0.904 | 0.194 | 0.269 | — | — | |
| DropGIN2024.02 | 0.17 | 0.55 | 0.44 | 0.11 | 0.12 | — | 29.87 | 0.03 | 0.03 | 29.82 | 24.74 | 0.38 | 0.4 | 0.38 | 0.34 | 0.97 | 0.19 | 0.27 | — | 0.37 | |
| TokenGT2024.02 | 0.177 | 0.755 | 0.447 | 0.123 | 0.13 | 31.54 | 0.03 | 15.477 | 12.449 | 11.418 | 29.721 | 0.18 | 0.304 | 0.302 | 0.312 | 0.273 | 3.823 | 0.109 | — | — | |
| TokenGT2024.02 | 0.18 | 0.76 | 0.45 | 0.12 | 0.13 | — | 31.54 | 0.03 | 0.03 | 15.48 | 11.42 | 0.18 | 0.3 | 0.31 | 0.27 | 3.82 | 0.11 | 0.1 | — | 0.3 | |
| GCN2024.02 | 0.184 | 0.629 | 0.525 | 0.128 | 0.136 | 33.913 | 0.035 | 30.971 | 25.276 | 30.924 | 26.138 | 0.221 | 0.371 | 0.39 | 0.396 | 0.979 | 0.295 | 0.284 | — | — | |
| BLIP-2 (LoRA)Category=VLM, requires LLM/VLM at inference=true, protocol=LoRA2026.05 | 4.92 | — | — | — | — | — | — | 22.12 | — | — | — | — | — | — | — | — | — | — | — | — | |
| GPT-4o (ICL)Category=LLM, requires LLM/VLM at inference=true, protocol=ICL2026.05 | 8.38 | — | — | — | — | — | — | 17.943 | — | — | — | — | — | — | — | — | — | — | — | — | |
| Janus-Pro 7B (ICL)Category=VLM, requires LLM/VLM at inference=true, protocol=ICL2026.05 | 8.53 | — | — | — | — | — | — | 17.211 | — | — | — | — | — | — | — | — | — | — | — | — | |
| EGD+EDMBase Diffusion Model=EDM2025.05 | 31 | 0.05 | 0.14 | 24 | 26 | — | — | — | — | — | — | 0.1 | — | — | — | — | — | — | — | — | |
| EGD+GeoLDMBase Diffusion Model=GeoLDM2025.05 | 33 | 0.03 | 0.88 | 15 | 30 | — | — | — | — | — | — | 0.13 | — | — | — | — | — | — | — | — | |
| EGD+TFGBase Diffusion Model=TFG2025.05 | 58 | 0.21 | 0.18 | 42 | 44 | — | — | — | — | — | — | 0.18 | — | — | — | — | — | — | — | — | |
| MUDM2025.05 | 85 | 0.33 | 0.43 | 72 | 133 | — | — | — | — | — | — | 0.29 | — | — | — | — | — | — | — | — | |
| TFG+TopNBudget=Same Evaluations (SE), Sampling Strategy=TopN2025.05 | 132 | 0.25 | 0.52 | 148 | 87 | — | — | — | — | — | — | 0.34 | — | — | — | — | — | — | — | — | |
| GeoLDM+TopNBudget=Same Evaluations (SE), Sampling Strategy=TopN2025.05 | 144 | 0.22 | 0.85 | 99 | 120 | — | — | — | — | — | — | 0.65 | — | — | — | — | — | — | — | — | |
| EDM+TopNBudget=Same Evaluations (SE), Sampling Strategy=TopN2025.05 | 174 | 0.26 | 0.89 | 97 | 129 | — | — | — | — | — | — | 0.1 | — | — | — | — | — | — | — | — | |
| GeoLDM+TopNBudget=Same Runtime (SR), Sampling Strategy=TopN2025.05 | 390 | 0.54 | 2.23 | 212 | 314 | — | — | — | — | — | — | 1.57 | — | — | — | — | — | — | — | — | |
| EDM+TopNBudget=Same Runtime (SR), Sampling Strategy=TopN2025.05 | 437 | 0.55 | 2.48 | 201 | 359 | — | — | — | — | — | — | 1.16 | — | — | — | — | — | — | — | — | |
| EEGSDE2025.05 | 487 | 0.78 | 2.5 | 302 | 447 | — | — | — | — | — | — | 0.94 | — | — | — | — | — | — | — | — | |
| TFG+TopNBudget=Same Runtime (SR), Sampling Strategy=TopN2025.05 | 550 | 0.74 | 2.81 | 322 | 407 | — | — | — | — | — | — | 1.84 | — | — | — | — | — | — | — | — | |
| cGeoLDM2025.05 | 587 | 1.11 | 2.37 | 340 | 522 | — | — | — | — | — | — | 1.03 | — | — | — | — | — | — | — | — | |
| cGCDM2025.05 | 595 | 0.86 | 1.99 | 346 | 480 | — | — | — | — | — | — | 0.7 | — | — | — | — | — | — | — | — | |
| cEDM2025.05 | 655 | 1.11 | 2.76 | 356 | 584 | — | — | — | — | — | — | 1.1 | — | — | — | — | — | — | — | — | |
| Atoms2025.05 | 866 | 1.05 | 3.86 | 426 | 813 | — | — | — | — | — | — | 1.97 | — | — | — | — | — | — | — | — | |
| TFG2025.05 | 893 | 1.33 | 3.9 | 984 | 568 | — | — | — | — | — | — | 2.77 | — | — | — | — | — | — | — | — | |
| Random2025.05 | 1,470 | 1.62 | 9.01 | 645 | 1,457 | — | — | — | — | — | — | 6.86 | — | — | — | — | — | — | — | — | |
| DropGINLearning Strategy=GW (No Transfer)2024.02 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 0.18 | — | |
| DropGINLearning Strategy=Inductive2024.02 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 0.159 | — | |
| DropGINLearning Strategy=Transductive2024.02 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 0.157 | — | |
| ESALearning Strategy=GW (No Transfer)2024.02 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 0.174 | — | |
| ESALearning Strategy=Inductive2024.02 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 0.15 | — | |
| ESALearning Strategy=Transductive2024.02 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 0.146 | — | |
| GATLearning Strategy=GW (No Transfer)2024.02 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 0.181 | — | |
| GATLearning Strategy=Inductive2024.02 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 0.156 | — | |
| GATLearning Strategy=Transductive2024.02 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 0.153 | — | |
| GATv2Learning Strategy=GW (No Transfer)2024.02 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 0.178 | — | |
| GATv2Learning Strategy=Inductive2024.02 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 0.157 | — | |
| GATv2Learning Strategy=Transductive2024.02 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 0.153 | — | |
| GCNLearning Strategy=GW (No Transfer)2024.02 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 0.181 | — | |
| GCNLearning Strategy=Inductive2024.02 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 0.161 | — | |
| GCNLearning Strategy=Transductive2024.02 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 0.159 | — | |
| GINLearning Strategy=GW (No Transfer)2024.02 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 0.18 | — | |
| GINLearning Strategy=Inductive2024.02 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 0.159 | — | |
| GINLearning Strategy=Transductive2024.02 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 0.155 | — | |
| GPSLearning Strategy=GW (No Transfer)2024.02 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 0.178 | — | |
| GPSLearning Strategy=Inductive2024.02 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 0.167 | — | |
| GPSLearning Strategy=Transductive2024.02 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 0.169 | — | |
| GraphormerLearning Strategy=GW (No Transfer)2024.02 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 0.19 | — | |
| GraphormerLearning Strategy=Inductive2024.02 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 0.151 | — | |
| GraphormerLearning Strategy=Transductive2024.02 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 0.147 | — | |
| PNALearning Strategy=GW (No Transfer)2024.02 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 0.174 | — | |
| PNALearning Strategy=Inductive2024.02 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 0.156 | — | |
| PNALearning Strategy=Transductive2024.02 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 0.153 | — | |
| TokenGTLearning Strategy=GW (No Transfer)2024.02 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 0.204 | — | |
| TokenGTLearning Strategy=Inductive2024.02 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 0.165 | — | |
| TokenGTLearning Strategy=Transductive2024.02 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 0.156 | — |