List your company

Reading Comprehension on RACE (3-shot)

83.31AccuracyBNRM
52.713260.656668.676.5434Feb 11, 2026

Evaluation results

MethodLinks
BNRMBase Model=Llama3.1-8B-Instruct, Evaluation Protocol=3-shot2026.0283.31
PoEBase Model=OpenRLHF-Llama3-8B-SFT, Evaluation Protocol=3-shot2026.0281.39
ALBMBase Model=OpenRLHF-Llama3-8B-SFT, Evaluation Protocol=3-shot2026.0280.69
LPBase Model=OpenRLHF-Llama3-8B-SFT, Evaluation Protocol=3-shot2026.0280.3
BaseBase Model=OpenRLHF-Llama3-8B-SFT, Evaluation Protocol=3-shot2026.0279.21
LPBase Model=Llama3.1-8B-Instruct, Evaluation Protocol=3-shot2026.0278.9
SKBase Model=OpenRLHF-Llama3-8B-SFT, Evaluation Protocol=3-shot2026.0278.82
InfoRMBase Model=OpenRLHF-Llama3-8B-SFT, Evaluation Protocol=3-shot2026.0278.72
BNRMBase Model=OpenRLHF-Llama3-8B-SFT, Evaluation Protocol=3-shot2026.0275.68
BaseBase Model=Llama3.1-8B-Instruct, Evaluation Protocol=3-shot2026.0266.5
InfoRMBase Model=Llama3.1-8B-Instruct, Evaluation Protocol=3-shot2026.0265.2
PoEBase Model=Llama3.1-8B-Instruct, Evaluation Protocol=3-shot2026.0260.03
ALBMBase Model=Llama3.1-8B-Instruct, Evaluation Protocol=3-shot2026.0259
SKBase Model=Llama3.1-8B-Instruct, Evaluation Protocol=3-shot2026.0253.89