TY  - RPRT
TI  - Multi-Turn Reinforcement Learning for Tool-Calling Agents with Iterative Reward Calibration
AU  - Wachiravit Modecrua
AU  - Krittanon Kaewtawee
AU  - Krittin Pachtrachai
AU  - Touchapon Kraisingkorn
PY  - 2026
UR  - https://arxiv.org/abs/2604.02869
ID  - 2604.02869
ER  -