Question Answering on HotpotQA (test) (Exact Match)
60Exact Match (EM)PersonalAI
Evaluation Results
| Method | Links | |
|---|---|---|
| PersonalAILLM=DeepSeek V3, Retrieval algorithm=hybrid of BeamSearch and WaterCircles, Graph restriction=thesis vertices excluded2025.06 | 60 | |
| PersonalAILLM=GPT4o mini, Retrieval algorithm=hybrid of BeamSearch and WaterCircles, Graph restriction=no restrictions applied2025.06 | 59 | |
| PersonalAILLM=Llama3.1 8B, Retrieval algorithm=BeamSearch, Graph restriction=object vertices excluded2025.06 | 37 | |
| PersonalAILLM=Qwen2.5 7B, Retrieval algorithm=A*, Graph restriction=no restrictions applied2025.06 | 18 | |
| PersonalAILLM=DeepSeek R1 7B, Retrieval algorithm=A*, Graph restriction=episodic vertices excluded2025.06 | 14 |