Question Answering on ESGenius
86.88AccuracyMOCA-AGENT
Evaluation Results
| Method | Links | |
|---|---|---|
| MOCA-AGENTSize=27B, IT (Instruction-tuned)=no, Reasoning-focused=no, RAG setting=true2026.06 | 86.88 | |
| Gemma-3Size=12B, IT (Instruction-tuned)=no, Reasoning-focused=yes, RAG setting=true2026.06 | 83.8 | |
| Gemma-3Size=27B, IT (Instruction-tuned)=no, Reasoning-focused=yes, RAG setting=true2026.06 | 83.36 | |
| Qwen2.5Size=72B, IT (Instruction-tuned)=no, Reasoning-focused=yes, RAG setting=true2026.06 | 82.57 | |
| Qwen2.5Size=32B, IT (Instruction-tuned)=no, Reasoning-focused=yes, RAG setting=true2026.06 | 82.47 | |
| Qwen2.5Size=14B, IT (Instruction-tuned)=no, Reasoning-focused=no, RAG setting=true2026.06 | 82.31 | |
| R1-Distill-LlamaSize=70B, IT (Instruction-tuned)=no, Reasoning-focused=yes, RAG setting=true2026.06 | 81.7 | |
| R1-Distill-QwenSize=32B, IT (Instruction-tuned)=no, Reasoning-focused=yes, RAG setting=true2026.06 | 81.43 | |
| Qwen2.5Size=32B, IT (Instruction-tuned)=no, Reasoning-focused=no, RAG setting=true2026.06 | 80.81 | |
| Qwen2.5Size=7B, IT (Instruction-tuned)=no, Reasoning-focused=no, RAG setting=true2026.06 | 80.55 | |
| R1-Distill-QwenSize=14B, IT (Instruction-tuned)=no, Reasoning-focused=yes, RAG setting=true2026.06 | 80.46 | |
| Llama-3.1Size=8B, IT (Instruction-tuned)=no, Reasoning-focused=yes, RAG setting=true2026.06 | 79.93 | |
| Qwen2.5Size=7B, IT (Instruction-tuned)=no, Reasoning-focused=yes, RAG setting=true2026.06 | 79.67 | |
| Qwen3Size=4B, IT (Instruction-tuned)=no, Reasoning-focused=no, RAG setting=true2026.06 | 79.05 | |
| Llama-3.3Size=70B, IT (Instruction-tuned)=no, Reasoning-focused=yes, RAG setting=true2026.06 | 78.87 | |
| Llama-3.1Size=8B, IT (Instruction-tuned)=no, Reasoning-focused=no, RAG setting=true2026.06 | 76.5 | |
| Qwen2.5Size=3B, IT (Instruction-tuned)=no, Reasoning-focused=no, RAG setting=true2026.06 | 76.32 | |
| QwQSize=32B, IT (Instruction-tuned)=no, Reasoning-focused=yes, RAG setting=true2026.06 | 76.14 | |
| Gemma-3Size=4B, IT (Instruction-tuned)=no, Reasoning-focused=yes, RAG setting=true2026.06 | 75.18 | |
| Llama-3Size=8B, IT (Instruction-tuned)=no, Reasoning-focused=no, RAG setting=true2026.06 | 73.24 | |
| Llama-3.2Size=3B, IT (Instruction-tuned)=no, Reasoning-focused=yes, RAG setting=true2026.06 | 72.18 | |
| Qwen2.5Size=72B, IT (Instruction-tuned)=no, Reasoning-focused=no, RAG setting=true2026.06 | 72.01 | |
| Qwen2.5Size=1.5B, IT (Instruction-tuned)=no, Reasoning-focused=yes, RAG setting=true2026.06 | 69.72 | |
| Qwen3Size=1.7B, IT (Instruction-tuned)=no, Reasoning-focused=no, RAG setting=true2026.06 | 69.37 | |
| Qwen2.5Size=1.5B, IT (Instruction-tuned)=no, Reasoning-focused=no, RAG setting=true2026.06 | 69.28 | |
| Gemma-3Size=4B, IT (Instruction-tuned)=no, Reasoning-focused=no, RAG setting=true2026.06 | 68.6 | |
| Gemma-3Size=12B, IT (Instruction-tuned)=no, Reasoning-focused=no, RAG setting=true2026.06 | 68.57 | |
| Llama-3.2Size=3B, IT (Instruction-tuned)=no, Reasoning-focused=no, RAG setting=true2026.06 | 68.31 | |
| Qwen3Size=8B, IT (Instruction-tuned)=no, Reasoning-focused=no, RAG setting=true2026.06 | 67.08 | |
| R1-Distill-QwenSize=7B, IT (Instruction-tuned)=no, Reasoning-focused=yes, RAG setting=true2026.06 | 65.05 | |
| Llama-3.2Size=1B, IT (Instruction-tuned)=no, Reasoning-focused=yes, RAG setting=true2026.06 | 64.52 | |
| R1-Distill-LlamaSize=8B, IT (Instruction-tuned)=no, Reasoning-focused=yes, RAG setting=true2026.06 | 62.5 | |
| Gemma-3Size=1B, IT (Instruction-tuned)=no, Reasoning-focused=yes, RAG setting=true2026.06 | 59.77 | |
| Qwen2.5Size=0.5B, IT (Instruction-tuned)=no, Reasoning-focused=no, RAG setting=true2026.06 | 53.96 | |
| Gemma-3Size=27B, IT (Instruction-tuned)=no, Reasoning-focused=no, RAG setting=true2026.06 | 52.29 | |
| R1-Distill-QwenSize=1.5B, IT (Instruction-tuned)=no, Reasoning-focused=yes, RAG setting=true2026.06 | 43.05 | |
| Llama-3.2Size=1B, IT (Instruction-tuned)=no, Reasoning-focused=no, RAG setting=true2026.06 | 36.8 | |
| Gemma-3Size=1B, IT (Instruction-tuned)=no, Reasoning-focused=no, RAG setting=true2026.06 | 25.26 |