Dejan Pecevski
Papers
1
Total Citations
54
H-Index
1
About
Dejan Pecevski is a computational neuroscientist whose research lies at the intersection of spiking neural networks, probabilistic inference, and cognitive computation. His most influential work, "Recurrent Spiking Networks Solve Planning Tasks" (2016, 54 citations), introduces a biologically plausible architecture that implements planning as probabilistic inference. By splitting the problem into stochastic transient firing dynamics and task-specific constraints, Pecevski demonstrates how recurrent spiking networks can solve finite and infinite horizon planning tasks—a significant step toward bridging neural dynamics with high-level cognition. This contribution highlights his broader focus on how networks of spiking neurons can perform complex computations traditionally reserved for symbolic AI. Pecevski’s work has been instrumental in advancing the understanding of neural coding and computation, earning recognition for its elegance and potential applications in neuromorphic engineering. His research continues to inspire students and researchers exploring the neural basis of decision-making, planning, and probabilistic reasoning.
Research Focus
Key Achievements
Top Papers
- 1Recurrent Spiking Networks Solve Planning Tasks54 citations · 2016