List your company

Deontic Reasoning on DeonticBench Airline (Accuracy)

36.3AccuracyGPT-5.2
-1.4528.34918.1527.951Jun 3, 2026

Evaluation results

MethodLinks
GPT-5.2Harness=terminus-kira2026.0636.3
Qwen3.5-397BHarness=direct2026.0619.2
GPT-5.2Harness=terminus-22026.0618.8
Qwen3.5-122BHarness=direct2026.0615
Qwen3.5-35BHarness=direct2026.0613.7
Qwen3-235BHarness=direct2026.0612.8
Qwen3.5-122BHarness=claude-code2026.0611.3
Qwen3.5-397BHarness=claude-code2026.0610
Kimi-K2Harness=direct2026.069
Qwen3.5-35BHarness=claude-code2026.068.8
Qwen3-Coder-480BHarness=claude-code2026.065
Qwen3.5-122BHarness=terminus-kira2026.065
Kimi-K2Harness=terminus-kira2026.065
Qwen3.5-122BHarness=terminus-22026.063.8
Qwen3-235BHarness=terminus-kira2026.063.8
GPT-5.2Harness=direct2026.062.5
Qwen3-Coder-480BHarness=direct2026.062.1
Qwen3-Coder-480BHarness=terminus-kira2026.061.3
Qwen3.5-122BHarness=codex2026.061.3
Qwen3.5-35BHarness=terminus-22026.061.3
Qwen3.5-35BHarness=terminus-kira2026.061.3
Qwen3.5-397BHarness=terminus-22026.061.3
Qwen3-235BHarness=terminus-22026.061.3
Qwen3-Coder-480BHarness=codex2026.060
Qwen3-Coder-480BHarness=terminus-22026.060
Qwen3.5-397BHarness=terminus-kira2026.060
Qwen3-235BHarness=codex2026.060
Kimi-K2Harness=codex2026.060
Kimi-K2Harness=terminus-22026.060
GPT-5.2Harness=codex2026.060