Or Avner

About

Or Avner is a researcher whose work bridges reinforcement learning, multi-agent systems, and goal-based decision-making. His most notable contribution is the introduction of **Sub-Goal Trees**, a framework that reimagines goal-based reinforcement learning (RL) by enabling agents to decompose complex tasks into manageable sub-goals. This approach, detailed in his 2020 paper (cited 8 times), addresses a fundamental limitation of traditional RL: its single-goal focus. By structuring learning around hierarchical sub-goals, Avner’s work allows agents to efficiently tackle multi-goal problems common in robotics and autonomous systems. His research has been instrumental in advancing how AI systems plan and execute long-horizon tasks, offering a more scalable alternative to flat reward optimization. Avner’s contributions are particularly impactful in domains requiring adaptive, goal-oriented behavior, such as robotic manipulation and navigation. With a growing citation footprint, his work continues to influence both theoretical RL and practical applications, positioning him as a key voice in the evolution of goal-conditioned and hierarchical reinforcement learning.

Research Focus

Key Achievements

Top Papers

1. 1
   Sub-Goal Trees -- a Framework for Goal-Based Reinforcement Learning

   8 citations · 2020

Key Collaborators

Contact & Links