List your company

Question Answering on HotpotQA In-distribution (test)

30.9AccuracyRLCR
17.43220.928524.42527.9215May 15, 2026

Evaluation results

MethodLinks
RLCRBackbone=Mistral-7B-Instruct2026.0530.90.203767.7
RLVRBackbone=Llama-3.1-8B-Instruct2026.0530.30.134981.18
RDBackbone=Mistral-7B-Instruct2026.0529.40.261572.79
CSRBackbone=Llama-3.1-8B-Instruct2026.0527.460.078388.86
CSRBackbone=Mistral-7B-Instruct2026.0527.320.088487.48
RLVRBackbone=Mistral-7B-Instruct2026.0527.30.131481.12
RLCRBackbone=Llama-3.1-8B-Instruct2026.0527.20.213766.16
RDBackbone=Llama-3.1-8B-Instruct2026.0526.40.11584.69
RLVRBackbone=Qwen2.5-7B-Instruct2026.0526.360.155980.83
RDBackbone=Qwen2.5-7B-Instruct2026.0525.50.139769.1
CSRBackbone=Qwen2.5-7B-Instruct2026.0524.10.115989.5
BaseBackbone=Qwen2.5-7B-Instruct2026.0523.990.155581.47
RLCRBackbone=Qwen2.5-7B-Instruct2026.0521.90.234773.12
BaseBackbone=Mistral-7B-Instruct2026.0518.370.160781.83
BaseBackbone=Llama-3.1-8B-Instruct2026.0517.950.182176.66