IUPAC and SMILES translation benchmark
56.5AccuracySTOUT
Evaluation results
| Method | Links | |
|---|---|---|
| STOUT2025.12 | 56.5 | |
| NatureLMNumber of parameters=8B2025.12 | 51.7 | |
| Qwen+MiST+SFTBase Model=Qwen-2.5 3B, Fine-tuning=SFT2025.12 | 44.5 | |
| LlaSMol_MistralBackbone=Mistral2025.12 | 29 | |
| NatureLMNumber of parameters=1B2025.12 | 28.4 | |
| GPT-42025.12 | 0 | |
| Claude 3 Opus2025.12 | 0 |