ResearchBenchmarksOut-of-domain Generalization on Diplomat, Mutual, Quality, CoQA, and Qasper (test)Follow70.9ScoreAutoMix-5.74814.15134.0553.949Oct 19, 2023Evaluation ResultsMethodMethodLinksScoreAutoMixSLM=GPT-3.5SLM=GPT-3.52023.1070.9AutoMixSLM=LLama-13bSLM=LLama-13b2023.1031.5AutoMixSLM=Mistral-7bSLM=Mistral-7b2023.1028.3FrugalGPTSLM=GPT-3.5SLM=GPT-3.52023.1014.3FrugalGPTSLM=Mistral-7bSLM=Mistral-7b2023.1012.5HybridLLMSLM=GPT-3.5SLM=GPT-3.52023.107.6HybridLLMSLM=Mistral-7bSLM=Mistral-7b2023.102.4FrugalGPTSLM=LLama-13bSLM=LLama-13b2023.100HybridLLMSLM=LLama-13bSLM=LLama-13b2023.10-2.8