Query Generation on Semantic Variants subset
98AccuracyGemini
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| GeminiMode=Zero-shot, Structured prompts=true2026.04 | 98 | 0 | 2 | 0 | |
| DeepSeek R1Mode=Zero-shot, Structured prompts=true2026.04 | 97.2 | 0 | 0.2 | 2.6 | |
| GPT4Mode=Zero-shot, Structured prompts=true2026.04 | 97.2 | 0 | 0.4 | 2.4 | |
| GrokMode=Zero-shot, Structured prompts=true2026.04 | 97 | 0 | 0 | 3 | |
| DeepSeek R1-Distill-8B fine-tuned with QLoRAMode=Fine-tuned2026.04 | 94.2 | 0 | 0 | 5.8 |