Multiple Choice Question Answering on IndoCulture native prompts (test)
67.5AccuracyGemma2-9B
Evaluation results
| Method | Links | |
|---|---|---|
| Gemma2-9BAugmentation=E-Path, Prompting=CCKG English Paths2026.01 | 67.5 | |
| Qwen2.5-7B-ITAugmentation=Native CCKG Paths2026.01 | 67.5 | |
| Qwen2.5-7B-ITAugmentation=None, Prompting=CoT2026.01 | 67.2 | |
| Qwen2.5-7B-ITAugmentation=English CCKG Paths2026.01 | 66.9 | |
| Qwen2.5-7B-ITAugmentation=Native CCKG Assertions2026.01 | 66.3 | |
| Qwen2.5-7B-ITAugmentation=Base2026.01 | 66.2 | |
| Qwen2.5-7B-ITAugmentation=English CCKG Assertions2026.01 | 66.1 | |
| Gemma2-9BAugmentation=Native CCKG Paths2026.01 | 65.9 | |
| Gemma2-9BAugmentation=Native CCKG Assertions2026.01 | 65.7 | |
| Qwen2.5-7B-ITAugmentation=Mango2026.01 | 65.3 | |
| Gemma2-9BAugmentation=Base2026.01 | 65.2 | |
| Gemma2-9BAugmentation=English CCKG Paths2026.01 | 65 | |
| Gemma2-9BAugmentation=Mango2026.01 | 64.8 | |
| Gemma2-9BAugmentation=English CCKG Assertions2026.01 | 64.6 | |
| Qwen2.5-7B-ITAugmentation=CoT2026.01 | 63.5 | |
| Gemma2-9B-ITAugmentation=Native CCKG Paths2026.01 | 61 | |
| Qwen2.5-7BAugmentation=Native CCKG Assertions2026.01 | 60.8 | |
| Qwen2.5-7BAugmentation=English CCKG Assertions2026.01 | 60.1 | |
| Qwen2.5-7BAugmentation=E-Path, Prompting=CCKG English Paths2026.01 | 59.8 | |
| Qwen2.5-7BAugmentation=Mango2026.01 | 59.7 | |
| Gemma2-9B-ITAugmentation=Native CCKG Assertions2026.01 | 59.4 | |
| Qwen2.5-7BAugmentation=English CCKG Paths2026.01 | 59.1 | |
| Gemma2-9B-ITAugmentation=English CCKG Paths2026.01 | 59.1 | |
| Llama3.1-8B-ITAugmentation=CoT2026.01 | 59 | |
| Qwen2.5-7BAugmentation=Native CCKG Paths2026.01 | 58.6 | |
| Gemma2-9B-ITAugmentation=N-Asrt, Prompting=CCKG Native Assertions2026.01 | 58.5 | |
| Qwen2.5-7BAugmentation=Base2026.01 | 58.5 | |
| Gemma2-9B-ITAugmentation=Base2026.01 | 57.9 | |
| Gemma2-9B-ITAugmentation=Mango2026.01 | 57.9 | |
| Gemma2-9B-ITAugmentation=English CCKG Assertions2026.01 | 57.3 | |
| Llama3.2-1BAugmentation=N-Path, Prompting=CCKG Native Paths2026.01 | 57.1 | |
| Llama3.1-8B-ITAugmentation=Native CCKG Paths2026.01 | 56.4 | |
| Llama3.1-8B-ITAugmentation=N-Path, Prompting=CCKG Native Paths2026.01 | 55.9 | |
| Llama3.1-8B-ITAugmentation=Base2026.01 | 55.5 | |
| Llama3.1-8B-ITAugmentation=English CCKG Paths2026.01 | 54.4 | |
| Llama3.1-8B-ITAugmentation=Native CCKG Assertions2026.01 | 54.4 | |
| Qwen2.5-3BAugmentation=Mango2026.01 | 54.3 | |
| Llama3.1-8B-ITAugmentation=English CCKG Assertions2026.01 | 54.2 | |
| Qwen2.5-3BAugmentation=Native CCKG Assertions2026.01 | 53.6 | |
| Qwen2.5-7BAugmentation=CoT2026.01 | 53.4 | |
| Llama3.2-1BAugmentation=Native CCKG Assertions2026.01 | 53.4 | |
| Llama3.1-8B-ITAugmentation=Mango2026.01 | 53.4 | |
| Qwen2.5-3BAugmentation=English CCKG Assertions2026.01 | 53.2 | |
| Gemma2-9BAugmentation=CoT2026.01 | 53.2 | |
| Qwen2.5-3BAugmentation=CoT2026.01 | 53.1 | |
| Llama3.2-1BAugmentation=Native CCKG Paths2026.01 | 53.1 | |
| Qwen2.5-3BAugmentation=Base2026.01 | 53 | |
| Qwen2.5-3BAugmentation=Native CCKG Paths2026.01 | 51.8 | |
| Qwen2.5-3BAugmentation=English CCKG Paths2026.01 | 51.7 | |
| Llama3.2-1BAugmentation=English CCKG Paths2026.01 | 51 | |
| Llama3.2-1BAugmentation=English CCKG Assertions2026.01 | 50.9 | |
| Llama3.2-3BAugmentation=Native CCKG Paths2026.01 | 50.5 | |
| Llama3.2-3BAugmentation=Native CCKG Assertions2026.01 | 50 | |
| Llama3.2-3BAugmentation=English CCKG Assertions2026.01 | 49.6 | |
| Llama3.2-3BAugmentation=Base2026.01 | 49.4 | |
| Llama3.2-3BAugmentation=English CCKG Paths2026.01 | 49.4 | |
| Llama3.2-3BAugmentation=Mango2026.01 | 49.2 | |
| Llama3.2-1BAugmentation=Mango2026.01 | 48.9 | |
| Llama3.2-3BAugmentation=N-Asrt, Prompting=CCKG Native Assertions2026.01 | 48.3 | |
| Llama3.2-1BAugmentation=Base2026.01 | 46.9 | |
| Qwen2.5-1.5BAugmentation=CoT2026.01 | 45.5 | |
| Qwen2.5-1.5BAugmentation=English CCKG Assertions2026.01 | 45.2 | |
| Qwen2.5-1.5BAugmentation=Mango2026.01 | 45 | |
| Qwen2.5-1.5BAugmentation=English CCKG Paths2026.01 | 44.8 | |
| Qwen2.5-1.5BAugmentation=Native CCKG Assertions2026.01 | 44.5 | |
| Llama3.2-1BAugmentation=CoT2026.01 | 44.5 | |
| Qwen2.5-1.5BAugmentation=Base2026.01 | 43.8 | |
| Qwen2.5-0.5BAugmentation=Mango2026.01 | 43.5 | |
| Qwen2.5-0.5BAugmentation=English CCKG Assertions2026.01 | 43.4 | |
| Qwen2.5-0.5BAugmentation=Base2026.01 | 43.1 | |
| Qwen2.5-0.5BAugmentation=English CCKG Paths2026.01 | 42.2 | |
| Qwen2.5-0.5BAugmentation=Native CCKG Paths2026.01 | 41.9 | |
| Qwen2.5-1.5BAugmentation=Native CCKG Paths2026.01 | 41.8 | |
| Qwen2.5-0.5BAugmentation=E-Asrt, Prompting=CCKG English Assertions2026.01 | 41.6 | |
| Qwen2.5-0.5BAugmentation=Native CCKG Assertions2026.01 | 41.6 | |
| Llama3.2-3BAugmentation=CoT2026.01 | 41.3 | |
| Qwen2.5-0.5BAugmentation=Mango, Prompting=CCKG2026.01 | 40.9 | |
| Qwen2.5-0.5BAugmentation=None, Prompting=Base2026.01 | 40.8 | |
| Qwen2.5-0.5BAugmentation=N-Asrt, Prompting=CCKG Native Assertions2026.01 | 40.6 | |
| Gemma2-2BAugmentation=Native CCKG Paths2026.01 | 39.4 | |
| Gemma2-9B-ITAugmentation=CoT2026.01 | 39.1 | |
| Qwen2.5-0.5BAugmentation=N-Path, Prompting=CCKG Native Paths2026.01 | 39 | |
| Qwen2.5-0.5BAugmentation=E-Path, Prompting=CCKG English Paths2026.01 | 38.5 | |
| Gemma2-2BAugmentation=Mango2026.01 | 38.3 | |
| Gemma2-2BAugmentation=English CCKG Paths2026.01 | 38.2 | |
| Gemma2-2BAugmentation=English CCKG Assertions2026.01 | 37.5 | |
| Qwen2.5-0.5BAugmentation=None, Prompting=CoT2026.01 | 37.3 | |
| Qwen2.5-0.5BAugmentation=CoT2026.01 | 35.8 | |
| Gemma2-2BAugmentation=Native CCKG Assertions2026.01 | 35.7 | |
| Llama3.1-8BAugmentation=None, Prompting=CoT2026.01 | 35.6 | |
| Llama3.1-8BAugmentation=CoT2026.01 | 35.6 | |
| Gemma2-2BAugmentation=CoT2026.01 | 35.5 | |
| Gemma2-2BAugmentation=Base2026.01 | 33.4 | |
| Llama3.1-8BAugmentation=Base2026.01 | 32.7 | |
| Llama3.1-8BAugmentation=Mango2026.01 | 32.7 | |
| Llama3.1-8BAugmentation=English CCKG Assertions2026.01 | 32.7 | |
| Llama3.1-8BAugmentation=English CCKG Paths2026.01 | 32.7 | |
| Llama3.1-8BAugmentation=Native CCKG Assertions2026.01 | 32.7 | |
| Llama3.1-8BAugmentation=Native CCKG Paths2026.01 | 32.7 |