List your company

Selective Generation on GSM8k (PRR)

64.2PRR (Accuracy)SATRMD+MSP
2.42418.46234.550.538Feb 20, 2025

Evaluation results

MethodLinks
SATRMD+MSPModel=Llama 8b v3.12025.0264.2
SAPLMAModel=Llama 8b v3.12025.0259.8
HUQ-SATRMDModel=Llama 8b v3.12025.0259.2
Lexical Similarity ROUGE-LModel=Llama 8b v3.12025.0246.7
SARModel=Llama 8b v3.12025.0245.5
EigenScoreModel=Llama 8b v3.12025.0243
Semantic EntropyModel=Llama 8b v3.12025.0242.4
Eccentricity NLI Score Entail.Model=Llama 8b v3.12025.0240.3
Maximum Sequence ProbabilityModel=Llama 8b v3.12025.0238
DegMat NLI Score Entail.Model=Llama 8b v3.12025.0235.7
EigValLaplacian NLI Score Entail.Model=Llama 8b v3.12025.0233.5
PerplexityModel=Llama 8b v3.12025.0225.9
SentenceSARModel=Llama 8b v3.12025.0215.1
FactoscopeModel=Llama 8b v3.12025.024.8