Qitong Gao

About

Qitong Gao is a researcher at the forefront of integrating formal methods with reinforcement learning for safe and reliable robot autonomy. His primary research areas span temporal logic-based motion planning, deep reinforcement learning, and decision-making under uncertainty. Gao’s most impactful contribution is the development of a model-free reinforcement learning method that synthesizes control policies for mobile robots to satisfy Linear Temporal Logic (LTL) specifications, even when transition probabilities are unknown—a work that has garnered 42 citations. He further advanced this line of inquiry by introducing Deep Imitative Q-learning (DIQL), which enables robots to navigate using noisy semantic observations while adhering to LTL constraints, addressing critical challenges in real-world perception. Most recently, Gao has tackled the limitations of Decision Transformers in stochastic environments by steering them via temporal difference learning, a notable achievement for offline reinforcement learning in robotics. His work bridges the gap between theoretical guarantees and practical deployment, making him a key figure in the quest for verifiably safe autonomous systems.

Research Focus

Key Achievements

Top Papers

1. 1
   Reduced variance deep reinforcement learning with temporal logic specifications

   42 citations · 2019
2. 2
   Deep Imitative Reinforcement Learning for Temporal Logic Robot Motion Planning with Noisy Semantic Observations

   3 citations · 2020
3. 3
   Steering Decision Transformers via Temporal Difference Learning

   2 citations · 2024

Key Collaborators

Contact & Links