Lingkai Kong's website
Postdoctoral Fellow at Harvard University
Welcome to Lingkai Kong (孔令恺)’s homepage! I am a postdoctoral fellow at Harvard, advised by Prof. Milind Tambe. I obtained my Ph.D. in Computational Science and Engineering from Georgia Institute of Technology, advised by Prof. Chao Zhang.
I develop reliable data-driven solutions for high-stakes and uncertain decision-making scenarios. My work is grounded in close collaboration with public sector partners in public health and environmental sustainability. Specifically, I work on the following areas:
1. Learning with uncertainty
Deep models can be confidently wrong, which is dangerous in high-stakes domains. I develop methods to quantify and calibrate uncertainty so models know when they don’t know and enable safer decisions.
- SDE-Net: Equipping Deep Neural Networks with Uncertainty Estimates, ICML’20
- Calibrated Language Model Fine-Tuning for In- and Out-of-Distribution Data, EMNLP’20
- When in Doubt: Neural Non-Parametric Uncertainty Quantification for Epidemic Forecasting, NeurIPS’21
- Two Birds with One Stone: Enhancing Uncertainty Quantification and Interpretability with Graph Functional Neural Process, AISTATS’24
2. Generative models for robust decision-making at scale
Forecasting is only the first step—predictions must ultimately inform decisions. For example, in conservation, rangers must strategically allocate patrols based on forecasts of poacher activity. However, uncertainty in these forecasts and the vast action space make decision-making especially challenging. I study how to leverage generative models to make optimization scalable and robust to uncertainty.
- Composite Flow Matching for Reinforcement Learning with Shifted-Dynamics Data, Arxiv
- Robust Optimization with Diffusion Models for Green Security, UAI’25
- Diffusion Models as Constrained Samplers for Optimization with Unknown Constraints, AISTATS’25
- AdaPlanner: Adaptive Planning from Feedback with Language Models, NeurIPS’23
- End-to-End Stochastic Optimization with Energy-Based Model, NeurIPS’22
3. Applications in social good
I work with NGOs and public agencies to translate methods into impact across public health, sustainability, and scientific discovery.
- Generative AI Against Poaching: Latent Composite Flow Matching for Wildlife Conservation, Arxiv
- Robust Optimization with Diffusion Models for Green Security, UAI’25
- LLM-based Agent Simulation for Maternal Health Interventions: Uncertainty Estimation and Decision-focused Evaluation, AAMAS-AASG’25
- PRIORITY2REWARD: Incorporating Healthworker Preferences for Resource Allocation Planning, AAAI’25
- When in Doubt: Neural Non-Parametric Uncertainty Quantification for Epidemic Forecasting, NeurIPS’21
- Efficient Evolutionary Search Over Chemical Space with Large Language Models, ICLR’25