Harnessing Uncertainty: Entropy-Modulated Policy Gradients for Long-Horizon LLM Agents Paper • 2509.09265 • Published Sep 11 • 45