abel-causal-benchmark-v2
by zeyuyuyu
Run Abel Causal Benchmark V2 with 35 forward-looking prediction questions. Features 10 intervention scenarios, CAP (Causal Agent Protocol) mapping, and Enhanced CEVS scoring. Tests "causal emotional value" through explainable predictions, what-if analysis, and multi-hop causal chains that surpass generic LLM capabilities.