Spectra-to-SMILES translation on SpecMol-Bench (test)
52.44Token AccuracySpecMol-7B
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| SpecMol-7BNormalization=Standardized numerical representations2025.09 | 52.44 | 27 | 55.55 | 47.62 | 56.74 | |
| SpecMol-7BNormalization=w.o. Norm2025.09 | 50.55 | 25.5 | 53.98 | 46.79 | 55.81 | |
| Gemini-3-flashModel Category=Generalist Models2025.09 | 25.58 | 2.01 | 30.91 | 23.88 | 34.12 | |
| KIMI-K2Model Category=Generalist Models2025.09 | 19.72 | 0.5 | 24.73 | 16.94 | 27.75 | |
| InternS1-235BModel Category=Specialist Models2025.09 | 18.84 | 0 | 26.59 | 18.45 | 28.27 | |
| GPT-5Model Category=Generalist Models2025.09 | 18.82 | 1 | 21.69 | 14.27 | 25.15 | |
| o3-miniModel Category=Generalist Models2025.09 | 18.12 | 1.5 | 22.25 | 14.59 | 24.33 | |
| Qwen3-235BModel Category=Generalist Models2025.09 | 16.51 | 0 | 21.27 | 13.47 | 23.02 | |
| Gemini-2.5-flashModel Category=Generalist Models2025.09 | 16.02 | 0 | 19.57 | 22.92 | 12.6 | |
| DeepSeek-V3-685BModel Category=Generalist Models2025.09 | 15.84 | 0 | 20.02 | 14.1 | 22.08 | |
| ChemDFM-v2.0-14BModel Category=Specialist Models2025.09 | 12.34 | 0 | 18.95 | 9.62 | 17.78 | |
| Qwen2.5-7BModel Category=Generalist Models, Model variant=Base2025.09 | 6.56 | 0 | 3.93 | 2.63 | 4.16 |