Proficiency-Level Control zero-shot Instruct
31F1 ScoreLlama3-8BF
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| Llama3-8BFFine-tuning=Ontology-Guided CLM, Prompting=Zero-shot, Variant=Instruct2026.04 | 31 | 0.15 | 19 | 1.22 | 72 | |
| Phi-3.5-miniFFine-tuning=Ontology-Guided CLM, Prompting=Zero-shot, Variant=Instruct2026.04 | 24 | 0.14 | 19 | 1.56 | 25 | |
| Mistral-7B-v0.3FFine-tuning=Ontology-Guided CLM, Prompting=Zero-shot, Variant=Instruct2026.04 | 24 | 0.2 | 25 | 1.57 | 41 | |
| Llama3.2-3BFFine-tuning=Ontology-Guided CLM, Prompting=Zero-shot, Variant=Instruct2026.04 | 23 | 0.14 | 23 | 1.48 | 64 | |
| DeepSeek-R1-8BFFine-tuning=Ontology-Guided CLM, Prompting=Zero-shot, Variant=Instruct2026.04 | 23 | 0.13 | 26 | 1.4 | 20 | |
| Llama3.1-8BFFine-tuning=Ontology-Guided CLM, Prompting=Zero-shot, Variant=Instruct2026.04 | 22 | 0.17 | 23 | 1.57 | 82 | |
| Qwen2.5-7BFFine-tuning=Ontology-Guided CLM, Prompting=Zero-shot, Variant=Instruct2026.04 | 20 | 0.14 | 20 | 1.77 | 94 | |
| Llama3.1-8BFine-tuning=Pre-Trained Baseline, Prompting=Zero-shot, Variant=Instruct2026.04 | 14 | 0.09 | 19 | 1.98 | — | |
| Qwen2.5-7BFine-tuning=Pre-Trained Baseline, Prompting=Zero-shot, Variant=Instruct2026.04 | 14 | 0.05 | 18 | 2.01 | — | |
| Mistral-7B-v0.3Fine-tuning=Pre-Trained Baseline, Prompting=Zero-shot, Variant=Instruct2026.04 | 14 | 0.06 | 15 | 2.18 | — | |
| DeepSeek-R1-8BFine-tuning=Pre-Trained Baseline, Prompting=Zero-shot, Variant=Instruct2026.04 | 14 | 0.03 | 14 | 1.65 | — | |
| Phi-3.5-miniFine-tuning=Pre-Trained Baseline, Prompting=Zero-shot, Variant=Instruct2026.04 | 13 | 0.04 | 16 | 2.13 | — | |
| Llama3.2-3BFine-tuning=Pre-Trained Baseline, Prompting=Zero-shot, Variant=Instruct2026.04 | 12 | 0.04 | 18 | 2.19 | — | |
| Llama3-8BFine-tuning=Pre-Trained Baseline, Prompting=Zero-shot, Variant=Instruct2026.04 | 6 | 0 | 16 | 2.42 | — |