List your company

Deontic Reasoning on DeonticBench USCIS-AAO

81.9Macro F1GPT-5.2
31.87644.86357.8570.837Jun 3, 2026

Evaluation results

MethodLinks
GPT-5.2Harness=codex2026.0681.9
GPT-5.2Harness=terminus-22026.0678.1
Qwen3.5-122BHarness=direct2026.0678
GPT-5.2Harness=direct2026.0677.9
Qwen3.5-397BHarness=terminus-kira2026.0677.8
Qwen3.5-122BHarness=codex2026.0677.5
Qwen3.5-122BHarness=terminus-kira2026.0676.4
Qwen3.5-122BHarness=claude-code2026.0673
Qwen3.5-397BHarness=direct2026.0672.7
Qwen3.5-35BHarness=terminus-kira2026.0671.8
GPT-5.2Harness=terminus-kira2026.0671.3
Qwen3.5-397BHarness=terminus-22026.0670.8
Qwen3-235BHarness=terminus-22026.0668.9
Qwen3-235BHarness=terminus-kira2026.0666.8
Kimi-K2Harness=terminus-kira2026.0666.8
Qwen3.5-397BHarness=claude-code2026.0664.3
Qwen3.5-35BHarness=terminus-22026.0660.7
Qwen3.5-122BHarness=terminus-22026.0660.3
Qwen3.5-35BHarness=claude-code2026.0660.3
Kimi-K2Harness=codex2026.0655.3
Kimi-K2Harness=terminus-22026.0653.3
Qwen3-235BHarness=direct2026.0653.1
Kimi-K2Harness=direct2026.0651.8
Qwen3-235BHarness=codex2026.0650.9
Qwen3-Coder-480BHarness=claude-code2026.0650.5
Qwen3.5-35BHarness=direct2026.0647.7
Qwen3-Coder-480BHarness=codex2026.0642.7
Qwen3-Coder-480BHarness=terminus-22026.0640.8
Qwen3-Coder-480BHarness=terminus-kira2026.0637.8
Qwen3-Coder-480BHarness=direct2026.0633.8