List your company

BDI/E State Understanding on BDI/E State Inference Comprehension Mode (test)

71.1E-A MetricCogWM
30.33240.91651.562.084Jun 28, 2026

Evaluation results

MethodLinks
CogWMModel Size=14B2026.0671.173.967.1443.712.923.274.29
GPT-5.52026.0642.348.945.747.12.392.052.191.92
Qwen3Model Size=14B2026.0632.439.836.141.92.331.631.791.55
DSv4-Pro2026.0631.946.84342.51.952.022.21.89