List your company

Multiple-choice Question Answering on MMLU-Pro (naive task stream)

91Accuracy (Economics)ExpRAG
85.887.1588.589.85Jun 1, 2026

Evaluation results

MethodLinks
ExpRAGPass=1st2026.0691—81.3—7380
ExpRAGPass=2nd2026.0691——837880
MEMPROBEPass=1st2026.0691—82.7—7681
MEMPROBEPass=2nd2026.0691——837682
ReActPass=1st2026.0690—78.7—6680
ReActPass=2nd2026.0690——78.76680
ReMemPass=2nd2026.0687——81.78177
ReMemPass=1st2026.0686—79—7477