ResearchBenchmarksHonesty Alignment on OOD Aggregate AverageFollow0.8447AUROCEliCal0.793220.8065850.819950.833315Oct 20, 2025Evaluation ResultsMethodMethodLinksAUROCEliCalModel=Qwen-7B, Regime=...Model=Qwen-7B, Regime=Training-based, Sample Size=1k2025.100.8447EliCalModel=Qwen-14B, Regime...Model=Qwen-14B, Regime=Training-based, Sample Size=1k2025.100.8279EliCalModel=Llama-8B, Regime...Model=Llama-8B, Regime=Training-based, Sample Size=1k2025.100.7952