UAlign (reaction class as prior)

Reported on 4 benchmarks across 1 task · 1 paper · 3 SOTA

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Medical4 results

Single-step retrosynthesisonUSPTO-50k
Top-10 accuracy· 2024-03-25
95
best: 96.9 (O-GNN (reaction class as prior))
SOTA
UAlign: Pushing the Limit of Template-free Retrosynthesis Prediction with Unsupervised SMILES Alignment arXiv:2404.00044
Single-step retrosynthesisonUSPTO-50k
Top-3 accuracy· 2024-03-25
86.7
best: 87.7 (O-GNN (reaction class as prior))
SOTA
UAlign: Pushing the Limit of Template-free Retrosynthesis Prediction with Unsupervised SMILES Alignment arXiv:2404.00044
Single-step retrosynthesisonUSPTO-50k
Top-5 accuracy· 2024-03-25
91.5
best: 93.4 (O-GNN (reaction class as prior))
SOTA
UAlign: Pushing the Limit of Template-free Retrosynthesis Prediction with Unsupervised SMILES Alignment arXiv:2404.00044
Single-step retrosynthesisonUSPTO-50k
Top-1 accuracy· 2024-03-25
66.4
best: 67.2 (NAG2G (reaction class as prior))
UAlign: Pushing the Limit of Template-free Retrosynthesis Prediction with Unsupervised SMILES Alignment arXiv:2404.00044