Bharat Prakash

Papers

Total Citations

H-Index

About

Bharat Prakash is an emerging researcher specializing in reinforcement learning, hierarchical agent architectures, and real-world robot navigation. His work addresses some of the most persistent challenges in applied AI: enabling autonomous agents to learn efficiently in complex environments characterized by sparse rewards, multiple goals, and extended task horizons. In his notable work "ReProHRL: Towards Multi-Goal Navigation in the Real World using Hierarchical Agents" (2023), Prakash investigates how hierarchical reinforcement learning frameworks can bridge the simulation-to-reality gap, allowing robots to transfer learned policies into real-world multi-goal scenarios. Complementing this, his 2021 paper "Interactive Hierarchical Guidance using Language" explores how natural language can be leveraged to decompose complex tasks and improve sample efficiency — a critical bottleneck in practical RL deployments. With a combined citation count reflecting a growing body of interest in his work, Prakash contributes to the frontier of making autonomous systems more adaptable and practical outside controlled settings. His research is particularly valuable for students and practitioners working at the intersection of robotics, natural language processing, and hierarchical decision-making, offering novel approaches to challenges that continue to limit real-world AI deployment.