Model openai4110

Training model: gpt-4.1
Inference model: gpt-4o-mini

Investigations

Performance

DatasetValidation accuracyTest accuracy
espionage0.90.9
potions0.850.45
southgermancredit0.61538461538461540.6106194690265486
timetravel_insurance0.90.75
titanic0.72549019607843130.6470588235294118
wisconsin0.88888888888888880.7878787878787878