List your company

Multi-task Language Understanding on MMLU-Redux (unseen categories)

72AccuracyAFlow
55.3659.686468.32Feb 23, 2026

Evaluation results

MethodLinks
AFlowBase LLM=GPT-5-Mini2026.0272
GDesignerBase LLM=GPT-5-Mini2026.0272
HieraMASBase LLM=Qwen3-80B2026.0272
SC+CoTBase LLM=GPT-5-Mini2026.0268
AFlowBase LLM=Qwen3-80B2026.0268
GDesignerBase LLM=Qwen3-80B2026.0268
MASRouterBase LLM=Qwen3-80B2026.0268
HieraMASBase LLM=GPT-5-Mini2026.0268
SC+CoTBase LLM=Qwen3-80B2026.0264
MASRouterBase LLM=GPT-5-Mini2026.0256