Uncertainty-Aware LLM-Guided Policy Shaping for Sparse-Reward Reinforcement Learning

Ujjwal Bhatta, Utsabi Dangol, Sumaly Bajracharya +2

Sparse rewards and heterogeneous task sequences remain persistent challenges in Reinforcement Learning (RL), often resulting in slow convergence, weak generalization, and inefficie…