ESCORT: Efficient Stein-variational and Sliced Consistency-Optimized Temporal Belief Representation for POMDPs
Proceedings of the Thirty-Ninth Annual Conference on Neural Information Processing Systems (NeurIPS), 2025
Summary: Develops particle-based belief representation using Stein-variational methods with correlation-aware projections to accurately model multi-modal distributions in POMDPs.
Citation: Zhang, Y., Luo, B., Mukhopadhyay, A., Karsai, G., & Dubey, A. (2025). ESCORT: Efficient Stein-variational and Sliced Consistency-Optimized Temporal Belief Representation for POMDPs. In Proceedings of NeurIPS 2025.
Observation Adaptation via Annealed Importance Resampling for Partially Observable Markov Decision Processes
Proceedings of the 35th International Conference on Automated Planning and Scheduling (ICAPS), 2025 - Oral Presentation
Summary: Proposes annealed importance resampling to address particle degeneracy in online POMDP solvers when observations are highly informative, improving belief approximation accuracy.
Citation: Zhang, Y., Luo, B., Mukhopadhyay, A., & Dubey, A. (2025). Observation Adaptation via Annealed Importance Resampling for Partially Observable Markov Decision Processes. In Proceedings of ICAPS 2025. AAAI Press.
In-Context Planning with Latent Temporal Abstractions
Under review at the Fourteenth International Conference on Learning Representations (ICLR 2026)
Summary: Unifies in-context adaptation and online planning in learned latent temporal-abstraction space, enabling efficient decision-making under stochastic dynamics and partial observability.
Citation: Luo, B., Zhang, Y., Keplinger, N.S., Gupta, S., Dubey, A., & Mukhopadhyay, A. (2025). In-Context Planning with Latent Temporal Abstractions. Under review at ICLR 2026.
Act as You Learn: Adaptive Decision-Making in Non-Stationary Markov Decision Processes
Proceedings of the 23rd International Conference on Autonomous Agents and Multiagent Systems (AAMAS), 2024
Summary: Introduces ADA-MCTS algorithm that adaptively balances pessimistic and optimistic planning based on epistemic and aleatoric uncertainty estimates in non-stationary environments.
Citation: Luo, B., Zhang, Y., Dubey, A., & Mukhopadhyay, A. (2024). Act as You Learn: Adaptive Decision-Making in Non-Stationary Markov Decision Processes. In Proceedings of AAMAS 2024 (pp. 1301-1309).
Decision Making in Non-Stationary Environments with Policy-Augmented Search
Proceedings of the 23rd International Conference on Autonomous Agents and Multiagent Systems (AAMAS), 2024
Summary: Introduces PA-MCTS algorithm that combines stale policy knowledge with online Monte Carlo tree search to handle non-stationary environments where relearning optimal policies is computationally expensive.
Citation: Pettet, A., Zhang, Y., Luo, B., Wray, K., Baier, H., Laszka, A., Dubey, A., & Mukhopadhyay, A. (2024). Decision Making in Non-Stationary Environments with Policy-Augmented Search. In Proceedings of AAMAS 2024 (pp. 2417-2419).
NS-Gym: Open-Source Simulation Environments and Benchmarks for Non-Stationary Markov Decision Processes
Proceedings of the Thirty-Ninth Annual Conference on Neural Information Processing Systems (NeurIPS), Datasets and Benchmarks Track, 2025
Summary: Presents first standardized toolkit for non-stationary MDPs integrated with Gymnasium framework, enabling reproducible evaluation of algorithms under changing environmental conditions.
Citation: Keplinger, N.S., Luo, B., Bektas, I., Zhang, Y., Wray, K.H., Laszka, A., Dubey, A., & Mukhopadhyay, A. (2025). NS-Gym: Open-Source Simulation Environments and Benchmarks for Non-Stationary Markov Decision Processes. In Proceedings of NeurIPS 2025, Datasets and Benchmarks Track.
Online Decision-Making Under Uncertainty for Vehicle-to-Building Systems
Proceedings of the ACM/IEEE 16th International Conference on Cyber-Physical Systems (ICCPS), 2025
Summary: Models vehicle-to-building energy optimization as an MDP and uses domain-guided Monte Carlo tree search to handle heterogeneous chargers and demand charges over long planning horizons.
Citation: Sen, R., Zhang, Y., Liu, F., Talusan, J.P., Pettet, A., Suzue, Y., Mukhopadhyay, A., & Dubey, A. (2025). Online Decision-Making Under Uncertainty for Vehicle-to-Building Systems. In Proceedings of ICCPS 2025.
Neuro-symbolic AI: Foundations and Applications (Book Chapter)
Wiley-IEEE Press, 2025
Summary: Contributed Chapter 4 on neuro-symbolic approaches for UAV navigation and search planning in urban rescue operations.
Citation: Velasquez, A., Song, H., Ravikumar, P., Sastry, S.S., & Neema, S. (Eds.). (2025). Neuro-symbolic AI: Foundations and Applications. Wiley-IEEE Press. ISBN: 978-1-394-30237-6.
Shrinking POMCP: A Framework for Real-Time UAV Search and Rescue
2024 International Conference on Assured Autonomy (ICAA), 2024
Summary: Presents Shrinking POMCP that reduces computational complexity in UAV search and rescue by dynamically guiding exploration toward non-sparse belief regions.
Citation: Zhang, Y., Luo, B., Mukhopadhyay, A., Stojcsics, D., Elenius, D., Roy, A., Jha, S., Maroti, M., Koutsoukos, X., Karsai, G., & Dubey, A. (2024). Shrinking POMCP: A Framework for Real-Time UAV Search and Rescue. In Proceedings of ICAA 2024 (pp. 48-57).