publications
* indicates equal contribution.
2026
- ICLR’26
Token-Efficient Long-Term Interest Sketching and Internalized Reasoning for LLM-based RecommendationIn The Fourteenth International Conference on Learning Representations, 2026
2025
- NeurIPS’25E2former: An efficient and equivariant transformer with linear-scaling tensor productsIn The Thirty-ninth Annual Conference on Neural Information Processing Systems, 2025
- TKDELarge language models are in-context molecule learnersIEEE Transactions on Knowledge and Data Engineering, 2025
2024
- ICML’25GNNs-to-MLPs by Teacher Injection and Dirichlet Energy DistillationarXiv preprint arXiv:2412.11180, 2024
2023
2022
- NeurocomputingInfer-AVAE: An attribute inference model based on adversarial variational autoencoderNeurocomputing, 2022
2021
- NeurocomputingPayment-Guard: Detecting fraudulent in-app purchases in iOS systemNeurocomputing, 2021