Automated Scoring on NGSS-aligned modeling - Item 2 (test)
85.2AccuracyCA-Uniform
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| CA-UniformBackbone=vit_base_patch16_224, Fine-tuning strategy=LoRA, Confidence-aware scoring mechanism=Uniform2026.06 | 85.2 | 0.789 | 91.4 | 71.3 | 74.5 | |
| CA-SelectiveBackbone=vit_base_patch16_224, Fine-tuning strategy=LoRA, Confidence-aware scoring mechanism=Selective2026.06 | 85.2 | 0.789 | 91.4 | 71.3 | 74.5 | |
| ViT + LoRABackbone=vit_base_patch16_224, Fine-tuning strategy=LoRA, Confidence-aware scoring mechanism=None2026.06 | 83.3 | 0.759 | 90.5 | 69.5 | 72.9 | |
| ViT (Frozen)Backbone=vit_base_patch16_224, Fine-tuning strategy=Frozen, Confidence-aware scoring mechanism=None2026.06 | 48.2 | 0.134 | 30.4 | 32.7 | 29.9 |