List your company

Knowledge Composition Selection on HotpotQA

100Precision @2BASE
2.34427.69753.0578.403Aug 28, 2025

Evaluation results

MethodLinks
BASEType=MAX2025.0810087.0191.857897.5285.33
KCSBase Model=BERT-base2025.0864.1855.4658.750.1264.3355.52
KCSBase Model=BERT-large2025.0863.2954.7157.949.0962.9754.37
CLSBase Model=BERT-base2025.0862.754.0557.2648.6262.4253.87
CLSBase Model=BERT-large2025.085849.5252.6444.8359.6949.3
RETRIEVALBase Model=BM252025.0854.4446.749.5740.7952.3445.18
RETRIEVALBase Model=BM25, Execution=+step by step2025.0854.0846.649.3740.7552.4445.2
RETRIEVALBase Model=bge-large-v1.5, Execution=+step by step2025.0853.7846.0348.942.5254.2946.98
RETRIEVALBase Model=bge-large-v1.52025.0853.5645.7948.6641.2752.7845.63
LLMBase Model=GPT-4, prompt_mode=zero-shot2025.0852.143.7346.8345.156.7649.47
LLMBase Model=DeepSeek-V3, prompt_mode=zero-shot2025.0849.9341.6644.7143.0853.9547.14
RETRIEVALBase Model=bge-m32025.0849.2941.9744.6837.9848.3541.9
RETRIEVALBase Model=bge-small2025.0848.8341.4944.3338.1948.8242.22
RETRIEVALBase Model=text-ada-0022025.0847.4940.1942.8837.0546.9140.76
SENTENCE GRAPHStrategy=CommonEntity2025.0846.3940.0342.433.1142.7436.79
KCSBase Model=RoBERTa-large2025.0845.4239.3241.5937.0447.741.11
SENTENCE GRAPHStrategy=Similarity2025.0840.9535.437.4627.3935.530.5
CLSBase Model=RoBERTa-large2025.082319.5220.8320.3325.9322.47
LLMBase Model=Qwen2.5-7B, prompt_mode=zero-shot2025.0819.2616.1217.2815.919.9217.4
LLMBase Model=Qwen2.5-14B, prompt_mode=zero-shot2025.0818.115.1816.2614.7718.4916.16
LLMBase Model=Llama3.2-3B, prompt_mode=zero-shot2025.089.628.088.658.1110.188.88
LLMBase Model=Llama3.1-8B, prompt_mode=zero-shot2025.087.215.956.416.818.457.42
BASEType=Random2025.086.15.025.416.137.66.67