Priyam Parashar

Contextual Change (United States), University of California San Diego, Meta (United States)

Papers

Total Citations

H-Index

About

Priyam Parashar is a researcher at the intersection of cognitive robotics, human-robot teaming, and autonomous learning. Their work centers on how robots can reason about their own thinking—a field known as metareasoning—to become more adaptive and efficient learners. Parashar’s most influential paper, “Analogy and metareasoning: Cognitive strategies for robot learning” (2020, 11 citations), introduces a framework that enables robots to transfer knowledge across tasks by drawing analogies, significantly reducing the need for retraining. Building on this, “Meta-reasoning in Assembly Robots” (2021, 6 citations) applies these strategies to industrial settings, demonstrating how robots can dynamically allocate computational resources during complex assembly tasks. Parashar also contributed a foundational taxonomy in “A Taxonomy for Characterizing Modes of Interactions in Goal-driven, Human-robot Teams” (2019, 6 citations), which systematically categorizes how humans and autonomous agents collaborate in shared-goal environments. Earlier work, “Adaptive Agents in Minecraft” (2017, 5 citations), pioneered a hybrid paradigm blending domain knowledge with reinforcement learning, a technique now widely used in game-based AI research. Parashar’s recent “Situated Instruction Following” (2024) continues to push boundaries in natural language understanding for robots. With a growing citation footprint, Parashar is shaping how robots learn, reason, and collaborate with humans in real-world settings.