LLM-as-a-Judge on DDI (test)
59.03EM (Δ)GPT-4o-Mini
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| GPT-4o-MiniLLM-Generator Responses=Average of all LLM-Generators (Jahan et al. 2024)2025.06 | 59.03 | 1.84 | |
| Gemini-FlashLLM-Generator Responses=Average of all LLM-Generators (Jahan et al. 2024)2025.06 | 47.12 | 2.11 | |
| Qwen-2.5-7B-InstructLLM-Generator Responses=Average of all LLM-Generators (Jahan et al. 2024)2025.06 | 46.6 | 2.15 | |
| Phi-3.5-Mini-3.8B-InstructLLM-Generator Responses=Average of all LLM-Generators (Jahan et al. 2024)2025.06 | 43.06 | 2.19 | |
| Deepseek-R1-Qwen-7BLLM-Generator Responses=Average of all LLM-Generators (Jahan et al. 2024)2025.06 | 42.67 | 3.07 | |
| Deepseek-R1-LLaMA-8BLLM-Generator Responses=Average of all LLM-Generators (Jahan et al. 2024)2025.06 | 42.15 | 4.16 | |
| Claude-3-HaikuLLM-Generator Responses=Average of all LLM-Generators (Jahan et al. 2024)2025.06 | 31.15 | 2.7 | |
| LLaMA-3.1-8B-InstructLLM-Generator Responses=Average of all LLM-Generators (Jahan et al. 2024)2025.06 | 29.32 | 2.95 |