Dynamic Interactive Conversation on GOODAILTM (120K)
6.7Solved ScenariosOBLIVION
Evaluation Results
| Method | Links | |
|---|---|---|
| OBLIVIONModel=GPT-4.1-mini2026.03 | 6.7 | |
| FULLCTXModel=GPT-4.1-mini, Type=direct2026.03 | 6.42 | |
| BEYONDPROMPTSModel=GPT-4.1-mini, Type=memory2026.03 | 6.26 | |
| OBLIVIONModel=GPT-4o-mini2026.03 | 5.63 | |
| OBLIVIONModel=Qwen3-30B-A3B†2026.03 | 5.3 | |
| BEYONDPROMPTSModel=GPT-4o-mini, Type=memory2026.03 | 4.93 | |
| FULLCTXModel=GPT-4o-mini, Type=direct2026.03 | 4.81 | |
| FULLCTXModel=Qwen3-30B-A3B†, Type=direct2026.03 | 4.25 | |
| BEYONDPROMPTSModel=Qwen3-30B-A3B†, Type=memory2026.03 | 4.17 | |
| OBLIVIONModel=Phi-4-mini†2026.03 | 3.37 | |
| BEYONDPROMPTSModel=Phi-4-mini†, Type=memory2026.03 | 2.52 | |
| FULLCTXModel=Phi-4-mini†, Type=direct2026.03 | 1.15 |