Papers
3
Total Citations
10
H-Index
2
About
Robert Wright is an emerging researcher at the forefront of reinforcement learning (RL) and intelligent agent systems, with a particular focus on making sequential decision-making more efficient and scalable. His work addresses one of the most pressing challenges in modern RL: the prohibitive cost of learning complex, multi-step tasks from scratch. Wright has made notable contributions across three interconnected areas — automaton-guided curriculum generation, few-shot policy transfer, and LLM-integrated task planning. His 2023 paper on automaton-guided curriculum generation proposes innovative methods for automatically structuring learning experiences using logical task specifications, helping agents navigate otherwise intractable environments. Complementing this, his work on few-shot policy transfer via observation mapping and behavior cloning tackles the sim-to-real gap in robotics, enabling agents to rapidly adapt knowledge across domains with minimal interaction cost. Perhaps most ambitiously, his LgTS framework leverages Large Language Models to dynamically generate sub-goals for RL agents, advancing human-AI collaborative planning without restrictive assumptions about data availability. Though early in his career — with his papers accumulating citations since 2023 — Wright's research sits at a timely intersection of symbolic reasoning, deep learning, and robotics, positioning him as a promising voice in next-generation autonomous agent research.
Research Focus
Key Achievements
Top Papers
- 1Automaton-Guided Curriculum Generation for Reinforcement Learning Agents4 citations · 2023
- 2
- 3