Neuro-Symbolic Prompt Tuning for Trustworthy Reasoning in Large Language Models. ijaies 2026, 3 (1).