My research builds reliable agents for sequential decision making and planning, with a focus on safety and robustness in multi agent settings. I encode goals using formal specifications such as temporal logic and integrate them into learning-based planners, including diffusion-based trajectory generation for controllable, constraint-aware behavior.
News
- (Nov 2025) Heading to San Diego for NeurIPS 2025. Say hi if you're attending!
- (May 2025) Prelim done, on to the final stretch!
- (February 2025) Paper in Learning Scoring Rules in Autonomous Driving Planning Systems accepted to RA-L! Congrats Zikang!
- (September 2024) Paper in Scaling Specification-Guided Multi-Agent Control accepted to CoRL 2024!
- (May 2024) Paper in Offline Reinforcement Learning worked on during my summer at JPMorgan Chase & Co. accepted to ICML 2024! It was a pleasure collaborating with Sujay and Alec.
- (January 2024) Paper in Differentiable Signal Temporal Logic (STL) applied to robotics accepted to ICRA 2024, congrats Zikang!
- (October 2023) Paper in MCMC thinning using Stein Discrepancy worked on during my summer at JPMorgan Chase & Co. accepted to SIAM SIMODS!
- (June 2022) 2 Papers accepted to ECML PKDD 2022 in Temporal Logic controllers for Multi-agent RL systems and Adversarial Attacks to RL controllers. 1 accepted to IROS 2022 in Safer RL for Navigation!
Research
- 2024 conferenceScaling Safe Multi-Agent Control for Signal Temporal Logic Specifications
Joe Eappen, Zikang Xiong, Dipam Patel, Aniket Bera and Suresh Jagannathan.
Annual Conference on Robot Learning (CoRL 2024), Safe-ROL Workshop @ CoRL 2024
[paper] [code] [arXiv] [website] [bibTex] - 2024 conferenceInformation-Directed Pessimism for Offline Reinforcement Learning
Alec Koppel, Sujay Bhatt, Jiacheng Guo, Joe Eappen, Mengdi Wang, and Sumitra Ganesh.
International Conference on Machine Learning (ICML 2024)
[paper] [code] [bibTex] - 2024 conferenceCo-learning Planning and Control Policies Constrained by Differentiable Logic Specifications
Zikang Xiong, Daniel Lawson, Joe Eappen, Ahmed H. Qureshi, and Suresh Jagannathan.
IEEE International Conference on Robotics and Automation (ICRA 2024)
[paper] [demos] [arXiv] [bibTex] - 2022 conferenceDistSPECTRL: Distributing Specifications in Multi-Agent Reinforcement Learning Systems
Joe Eappen and Suresh Jagannathan.
European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases (ECML PKDD 2022), ALA Workshop @ AAMAS 2022
[paper] [code] [arXiv] [bibTex] - 2022 conferenceDefending Observation Attacks in Deep Reinforcement Learning via Detection and Denoising
Zikang Xiong, Joe Eappen, He Zhu, and Suresh Jagannathan.
European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases (ECML PKDD 2022)
[paper] [arXiv] [bibTex] - 2022 conferenceModel-free Neural Lyapunov Control for Safe Robot Navigation
Zikang Xiong, Joe Eappen, Ahmed H. Qureshi, and Suresh Jagannathan.
IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2022)
[paper] [demos] [arXiv] [bibTex]
Patents
- 2023 patentSystem and method for providing information-directed pessimism for offline reinforcement learning
Alec Koppel, Sujay Bhatt, Joe Eappen, and Sumitra Ganesh
US Patent Application No. 18/379,406, filed October 2023, published as US20250124334A1 [link]