Lingkai Kong

🎉 Excited to share: LSFLOW is accepted as a Spotlight at #ICML2026 (top 2%)!

The first flow/diffusion policy framework for RL with combinatorial actions. 🚀

📄 arxiv.org/abs/2601.22211
💻 github.com/anagha-satish/…

🧵👇

🎓 Thrilled to share: I'll be joining The University of Hong Kong, School of Computing and Data Science as a Tenure-track Assistant Professor in Fall 2026! We are building GenAI agents (diffusion / LLMs + RL) for real societal impact.

📣 Recruiting PhDs, a postdoc, and RAs — email lingkaikong@g.harvard.edu 🚀

#AcademicJobs #GenerativeAI #AgenticAI #LLM

🚀Excited to share that I will be attending @NeurIPSConf in San Diego next week!

NeurIPS2025d to present two works this year, both selected as Spotlights. These papers explore the intersection of Generative AI and decision-making, with a focus on sample efficiency and wildlife conservation applications.

Here are the details:

1️⃣ [𝗠𝗮𝗶𝗻 𝗧𝗿𝗮𝗰𝗸 𝗦𝗽𝗼𝘁𝗹𝗶𝗴𝗵𝘁] Composite Flow Matching for Reinforcement Learning. We introduce a method that leverages flow matching to extract prior knowledge from offline data to improve RL sample efficiency. By exploiting the connection to optimal transport, our method selectively incorporates useful priors even when facing distributional shifts in environmental dynamics.

📅 Time: Wed, Dec 3 | 4:30 PM - 7:30 PM PST

📍 Location: Exhibit Hall C, D, E #214

📄 Paper: arxiv.org/abs/2505.23062

💻 Code: github.com/Haichuan23/Com…

🤝 Collaborators: @HaichuanWang23 @tonghanwang Guojun Xiong @MilindTambe_AI

2️⃣ [𝗪𝗼𝗿𝗸𝘀𝗵𝗼𝗽 𝗦𝗽𝗼𝘁𝗹𝗶𝗴𝗵𝘁] Saving Wildlife with Generative AI: Latent Composite Flow Matching for Poaching Prediction at the Tackling Climate Change with Machine Learning Workshop. We propose the first Generative AI framework for poaching prediction. To address imperfect detection and data scarcity in conservation, we integrate flow matching with ecological occupancy models, injecting domain knowledge for better prediction.

📅 Time: Sun, Dec 7

📍 Location: Upper Level Room 24ABC

📄 Paper: arxiv.org/abs/2508.14342

🤝 Collaborators: @HaichuanWang23 @CEmogor Vincent Börsch-Supan @lilyxu0 @MilindTambe_AI

Happy to chat about GenAI for decision-making and applications in environmental science and public health. If you are around, let’s connect!

#NeurIPS2025 #GenerativeAI #ReinforcementLearning #AI4Good #DiffusionModels #FlowMatching #MachineLearning

Pre-collected offline data can supercharge RL—but what if the environment shifts and your transition dynamics no longer match?

Thrilled to share our NeurIPS 2025 Spotlight paper:
“Composite Flow Matching for RL with Shifted Dynamics Data”.

Paper: arxiv.org/abs/2505.23062
Code: github.com/Haichuan23/Com…

🧵1/
🚨 Poaching is a major threat to biodiversity.
Can we predict where it might strike next?
Meet Wildflow — our new latent flow matching framework for poaching prediction in wildlife conservation. 🐘🌱

📝 Preprint: arxiv.org/pdf/2508.14342