List your company

Mathematical Reasoning on GSM8K (test) (Acc, Latency)

90.52AccuracyOracle
66.308872.594478.8885.1656Aug 16, 2024

Evaluation results

MethodLinks
Oracle2024.0890.523.24
SELECTLLMPolicy=MLC + WEIGHTEDMAXCONF, s (number of LLMs)=42024.0877.9416.5
Top-s LLMss (number of LLMs)=42024.0877.4819
SELECTLLMPolicy=MLC + MAXCONF, s (number of LLMs)=42024.0877.4816.5
All LLMs2024.0876.0419
SELECTLLMPolicy=MLC + LABELLEDMAXCONF, s (number of LLMs)=42024.0875.6614.69
LLM-Blender2024.0875.2819
Random2024.0869.499.65
SELECTLLMPolicy=MLC + WEIGHTEDMAXCONF, s (number of LLMs)=12024.0867.244.7