ResearchBenchmarksConditional-independence (belief sufficiency) testing on Fever Expert-Constructed Bayesian NetworkFollow32.89CMILlama8.50214.833521.16527.4965Feb 6, 2026Evaluation ResultsMethodMethodLinksCMICMI 95% CILogLoss Imp. (CB: A~p vs A~(p,θ))LogLoss Imp. CI (CB: A~p vs A~(p,θ))LogLoss Imp. (CB: A~(p,x) vs A~(p,x,θ))LogLoss Imp. CI (CB: A~(p,x) vs A~(p,x,θ))LlamaModel=LlamaModel=Llama2026.0232.89—8.38—4.75—DeepSeekModel=DeepSeekModel=DeepSeek2026.0220.6—22.83—15.83—GPT-MinModel=GPT-MinModel=GPT-Min2026.0214.46—12.67—9.45—GPT-HighModel=GPT-HighModel=GPT-High2026.029.44—4.74—2.27—