All tasks combined on AOP-Wiki structured question answering set (test)
100AccuracyDeepSeek
Evaluation Results
| Method | Links | |
|---|---|---|
| DeepSeekRAG enhancement=W/ RAG2026.04 | 100 | |
| OverallRAG enhancement=W/ RAG2026.04 | 96.67 | |
| GeminiRAG enhancement=W/ RAG2026.04 | 95 | |
| GPTRAG enhancement=W/ RAG2026.04 | 95 | |
| DeepSeekRAG enhancement=W/O RAG2026.04 | 35 | |
| OverallRAG enhancement=W/O RAG2026.04 | 23.33 | |
| GeminiRAG enhancement=W/O RAG2026.04 | 20 | |
| GPTRAG enhancement=W/O RAG2026.04 | 15 |