Shibhansh Dohare

Papers

Total Citations

H-Index

About

Shibhansh Dohare is a rising researcher in artificial intelligence, focusing on reinforcement learning and temporal abstraction. His work addresses fundamental challenges in enabling agents to make decisions over long time horizons and to filter relevant information from noisy environments. Dohare’s most cited paper, “Gamma-Nets: Generalizing Value Estimation over Timescale” (2020, 12 citations), introduces a novel approach to temporal abstraction by generalizing value estimation across multiple timescales, a critical requirement for agents operating in complex, long-horizon tasks. This work has been influential in advancing multi-timescale value learning, providing a foundation for more efficient decision-making in reinforcement learning. More recently, in “Automatic Noise Filtering with Dynamic Sparse Training in Deep Reinforcement Learning” (2023, 2 citations), Dohare tackles the practical challenge of distinguishing useful information from noise, a key issue for robots performing diverse tasks in real-world settings. By integrating dynamic sparse training, this work offers a promising method for automatic noise filtering, enhancing the robustness and efficiency of deep reinforcement learning agents. Dohare’s contributions are shaping the future of autonomous systems, making his research highly relevant for students and practitioners aiming to build more adaptive and intelligent agents.