MPhil in Mathematics - A Sufficient Condition for Sample-Efficient Reinforcement Learning with General Function Approximation