#natural language process...
Apr 2026
Debating the Unspoken: Role-Anchored Multi-Agent Reasoning for Half-Truth Detection
Experiments show that RADAR consistently outperforms strong single- and multi-agent baselines across datasets and backbones, improving omission detection accuracy while reducing reasoning cost.
Yi-Xuan Tang, Yi-Rui Zhang, Hang Feng et al.
· arXiv.org · 1 citation