Uncovering Hidden Correctness in LLM Causal Reasoning via Symbolic Verification has been accepted to EACL (main)!