Binary Outcome Discrimination on Resolved vs. Rest (test)
90.1AUCQwen2.5-Coder-7B
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Qwen2.5-Coder-7BN=24,300, Res%=40.3%, index_formula=rho = Psi_state (endpoint-only)2026.06 | 90.1 | 79.4 | |
| Qwen2.5-Coder-14BN=24,300, Res%=49.6%, index_formula=rho = Psi_state (endpoint-only)2026.06 | 88.3 | 82.7 | |
| Qwen3-4B-BaseN=24,300, Res%=39.7%, index_formula=rho = Psi_state (endpoint-only)2026.06 | 88.2 | 74.8 | |
| Qwen3-8B-BaseN=24,300, Res%=43.7%, index_formula=rho = Psi_state (endpoint-only)2026.06 | 87.7 | 79.5 | |
| Qwen2.5-Coder-1.5BN=24,300, Res%=24.3%, index_formula=rho = Psi_state (endpoint-only)2026.06 | 87.5 | 57.7 | |
| CodeLlama-7BN=24,300, Res%=23.0%, index_formula=rho = Psi_state (endpoint-only)2026.06 | 87.5 | 58.6 | |
| Qwen2.5-7BN=24,300, Res%=35.3%, index_formula=rho = Psi_state (endpoint-only)2026.06 | 86.6 | 67.3 | |
| Qwen2.5-Coder-3BN=24,300, Res%=34.8%, index_formula=rho = Psi_state (endpoint-only)2026.06 | 86.3 | 70.2 | |
| Qwen3-1.7B-BaseN=24,300, Res%=25.8%, index_formula=rho = Psi_state (endpoint-only)2026.06 | 86.3 | 61.3 | |
| Qwen2.5-1.5BN=24,300, Res%=19.0%, index_formula=rho = Psi_state (endpoint-only)2026.06 | 86.2 | 55.7 | |
| AllN=372,600, Res%=29.2%, index_formula=rho = Psi_state (endpoint-only)2026.06 | 85 | 61.9 | |
| Qwen2.5-Coder-0.5BN=24,300, Res%=17.0%, index_formula=rho = Psi_state (endpoint-only)2026.06 | 84.7 | 46.4 | |
| DeepSeek-Coder-6.7BN=24,300, Res%=37.2%, index_formula=rho = Psi_state (endpoint-only)2026.06 | 84.1 | 71.2 | |
| Llama-2-7BN=24,300, Res%=20.3%, index_formula=rho = Psi_state (endpoint-only)2026.06 | 83.8 | 50.4 | |
| Qwen3-0.6B-BaseN=24,300, Res%=23.6%, index_formula=rho = Psi_state (endpoint-only)2026.06 | 80.8 | 52.2 | |
| Qwen2.5-0.5BN=24,300, Res%=17.9%, index_formula=rho = Psi_state (endpoint-only)2026.06 | 77.1 | 45 | |
| Qwen2.5-3BN=24,300, Res%=29.3%, index_formula=rho = Psi_state (endpoint-only)2026.06 | 76.4 | 55.3 |