List your company

Accuracy Evaluation on ROCStories (Writing)

76.3AccuracyES + TT-MV
16.81232.25647.763.144Mar 12, 2026

Evaluation results

MethodLinks
ES + TT-MVModel=OLMo3-7B2026.0376.3
ESModel=Llama3.1-8B-Inst2026.0368.1
ES + TT-MVModel=OLMo3-7B-Inst2026.0366
GRPOModel=OLMo3-7B-Inst2026.0365.8
ESModel=OLMo3-7B-Inst2026.0365.7
ES + TT-MVModel=Llama3.1-8B-Inst2026.0365.2
PPOModel=OLMo3-7B2026.0364.8
PPOModel=OLMo3-7B-Inst2026.0364.7
ESModel=Qwen2.5-3B-Inst2026.0364.5
RandOptModel=OLMo3-7B-Inst2026.0364.5
RandOptModel=OLMo3-7B2026.0364.5
ESModel=OLMo3-7B2026.0364.4
BaseModel=OLMo3-7B-Inst2026.0364
Best-of-N‡Model=OLMo3-7B-Inst2026.0363.5
TT-MV†Model=OLMo3-7B-Inst2026.0363.2
GRPOModel=OLMo3-7B2026.0362.2
GRPOModel=Llama3.1-8B-Inst2026.0362.2
ESModel=Qwen2.5-1.5B-Inst2026.0360.2
ES + TT-MVModel=Qwen2.5-1.5B-Inst2026.0359.1
RandOptModel=Llama3.1-8B-Inst2026.0359
PPOModel=Llama3.1-8B-Inst2026.0357.8
TT-MV†Model=Qwen2.5-3B-Inst2026.0357.3
RandOptModel=Qwen2.5-3B-Inst2026.0356.5
GRPOModel=Qwen2.5-3B-Inst2026.0356.3
Best-of-N‡Model=Qwen2.5-3B-Inst2026.0355
ES + TT-MVModel=Qwen2.5-3B-Inst2026.0355
BaseModel=Qwen2.5-3B-Inst2026.0354.7
Best-of-N‡Model=Llama3.1-8B-Inst2026.0354
GRPOModel=Qwen2.5-1.5B-Inst2026.0353.6
TT-MV†Model=Llama3.1-8B-Inst2026.0353.2
PPOModel=Qwen2.5-1.5B-Inst2026.0351.8
BaseModel=Llama3.1-8B-Inst2026.0351.8
Best-of-N‡Model=OLMo3-7B2026.0350
PPOModel=Qwen2.5-3B-Inst2026.0349
RandOptModel=Qwen2.5-1.5B-Inst2026.0348.5
BaseModel=Qwen2.5-1.5B-Inst2026.0346.7
TT-MV†Model=Qwen2.5-1.5B-Inst2026.0343.5
Best-of-N‡Model=Qwen2.5-1.5B-Inst2026.0342.2
TT-MV†Model=OLMo3-7B2026.0342.2
RandOptModel=Qwen2.5-0.5B-Inst2026.0332.2
ESModel=Qwen2.5-0.5B-Inst2026.0332
GRPOModel=Qwen2.5-0.5B-Inst2026.0330.9
ES + TT-MVModel=Qwen2.5-0.5B-Inst2026.0330.5
BaseModel=OLMo3-7B2026.0324.1
BaseModel=Qwen2.5-0.5B-Inst2026.0322
Best-of-N‡Model=Qwen2.5-0.5B-Inst2026.0320.3
TT-MV†Model=Qwen2.5-0.5B-Inst2026.0319.5
PPOModel=Qwen2.5-0.5B-Inst2026.0319.1