Multiple-choice question answering on Multiple-choice benchmarks GL
66.27AccuracyQwen3 14B merge-GL
Evaluation results
| Method | Links | |
|---|---|---|
| Qwen3 14B merge-GLBackbone=Qwen3 14B, Variant=merge-GL2026.03 | 66.27 | |
| ALIA 40B InstructBackbone=ALIA 40B, Variant=Instruct2026.03 | 64.98 | |
| Qwen3 14B merge-multiBackbone=Qwen3 14B, Variant=merge-multi2026.03 | 64.15 | |
| Qwen3 14B merge-ESBackbone=Qwen3 14B, Variant=merge-ES2026.03 | 64.02 | |
| Llama 3.1 8B merge-GLBackbone=Llama 3.1 8B, Variant=merge-GL2026.03 | 63.94 | |
| Qwen3 14B merge-CABackbone=Qwen3 14B, Variant=merge-CA2026.03 | 63.28 | |
| Qwen3 14B merge-EUBackbone=Qwen3 14B, Variant=merge-EU2026.03 | 62.66 | |
| Qwen3 14B InstructBackbone=Qwen3 14B, Variant=Instruct2026.03 | 62.39 | |
| Llama 3.1 8B merge-multiBackbone=Llama 3.1 8B, Variant=merge-multi2026.03 | 62.23 | |
| Llama 3.1 8B merge-EUBackbone=Llama 3.1 8B, Variant=merge-EU2026.03 | 61.41 | |
| Llama 3.1 8B merge-ESBackbone=Llama 3.1 8B, Variant=merge-ES2026.03 | 60.26 | |
| Llama 3.1 8B InstructBackbone=Llama 3.1 8B, Variant=Instruct2026.03 | 60.11 | |
| Llama 3.1 8B merge-CABackbone=Llama 3.1 8B, Variant=merge-CA2026.03 | 59.84 | |
| Qwen3 8B merge-GLBackbone=Qwen3 8B, Variant=merge-GL2026.03 | 58.84 | |
| Llama 3.1 8B joint-EUBackbone=Llama 3.1 8B, Variant=joint-EU2026.03 | 58.13 | |
| Qwen3 8B merge-ESBackbone=Qwen3 8B, Variant=merge-ES2026.03 | 56.74 | |
| Qwen3 8B merge-multiBackbone=Qwen3 8B, Variant=merge-multi2026.03 | 56.64 | |
| Qwen3 8B InstructBackbone=Qwen3 8B, Variant=Instruct2026.03 | 56.51 | |
| Qwen3 8B merge-CABackbone=Qwen3 8B, Variant=merge-CA2026.03 | 56.27 | |
| Qwen3 8B merge-EUBackbone=Qwen3 8B, Variant=merge-EU2026.03 | 55.8 | |
| Salamandra 7B InstructBackbone=Salamandra 7B, Variant=Instruct2026.03 | 53.6 | |
| Salamandra 2B InstructBackbone=Salamandra 2B, Variant=Instruct2026.03 | 37.11 |