Publication List of Midori Lab @ NTU (started 2024 August, at a certain pavillion in NTU, under the moonlight-)

Refer to our more digestible research highlights here.

P.s. we care more about works produced by our students and staff, than "corresponding author" shenanigans (unless some funding requirement states it)

**Denotes midori lab's direct research staff (e.g., postdoc)

*Denotes PhD or research students (supervised or co-supervised, or work done during visiting our lab)

Conferences

1. Gong, Z.*, Hou, Y.*, Zeng, Z.**, Xiao, M., Yuen, C., & Lim, W. Y. B. (2026). State of Thought Enables Endogenous Reasoning. In Advances in Neural Information Processing Systems (NeurIPS).

2. Guan, G.*, Hao, Y.**, Zhang, J.**, Zhang, F.*, & Lim, W. Y. B. (2026). C3H: Compression-to-Consensus Criteria Hijacking in Multimodal LLM Recommender Systems. In Advances in Neural Information Processing Systems (NeurIPS).

3. Zhang, J.**, Zhang, F.*, Xie, P., & Lim, W. Y. B. (2026). Natural Jammers: When Non-Robust Features Become Antagonistic. In Advances in Neural Information Processing Systems (NeurIPS) Spotlight.

4. Duan, Y.*, Wu, T.**, Hao, Y.**, & Lim, W. Y. B. (2026). WURI: Watching Unfolding Risk in Agent Interactions. In Advances in Neural Information Processing Systems (NeurIPS).

5. Delattre, B., Wu, H., Lim, W. Y. B., & Cao, Y. (2026). Joint Certification for Attributed Graphs: Beyond Topology-Only Robustness. In Advances in Neural Information Processing Systems (NeurIPS).

6. Zhang, J.**, Li, Z.*, Guo, S.*, Dong, J., Wang, K., & Lim, W. Y. B. (2026). FusionAudit: Pathway-Conditioned Robustness Auditing for Native Multimodal Models. In Advances in Neural Information Processing Systems (NeurIPS).

7. Chen, T.*, Hou, W., Wang, F., Wang, T.*, Zheng, Z., Tang, S., & Lim, W. Y. B. (2026). Stability-Enhanced Federated Learning with Accelerated Gradient. In Advances in Neural Information Processing Systems (NeurIPS).

8. Qiao, P., Cao, Y., Zhang, L., Cheng, G., Chen, J., Yu, M., Lim, W. Y. B., & Yoshikawa, M. (2026). Align Before Aggregation: Basis-Consistent Federated LoRA under Heterogeneous Ranks. In Advances in Neural Information Processing Systems (NeurIPS).

9. Hao, Y.**, Zhou, W.*, Guan, G.*, Wu, T.**, Zhang, F.*, & Lim, W. Y. B. (2027). VirusCascade: Hijacking Collaborative Reflection in LLM-Powered Recommender Agents. In Proceedings of the Network and Distributed System Security Symposium (NDSS).

10. Wang, C.*, Zhang, F.*, Zhang, J.**, Zhang, Z., Wang*, Y., Huang, L., Gao, J., Chen, Z., & Lim, W. Y. B. (2026). INSURE: Inference-Time Defense Framework for Securing Agents against Hijacking. In Proceedings of the ACM International Conference on Multimedia (ACM MM). ACM.

11. Zhao, Y.*, Zhang, W., Xiao, L., Zheng, Y., Liu, M.*, & Lim, W. Y. B. (2026). Advancing ESG Intelligence: An Expert-level Agent and Comprehensive Benchmark for Sustainable Finance. In Proceedings of the Conference on Empirical Methods in Natural Language Processing (EMNLP). Association for Computational Linguistics.

12. Wu, Y., Jin, H., Lim, W. Y. B., & Li, L. (2026). Perceive, Localize, Reflect: Evidence-Anchored Multi-Agent Debate for MLLM Hallucination Mitigation. In Proceedings of the Conference on Empirical Methods in Natural Language Processing (EMNLP). Association for Computational Linguistics.

13. Wang, C., Zhang, J.**, Zhang, Z., Wang, Z.*, Wang, Y., Gao, J., Wei, T., Chen, Z., & Lim, W. Y. B. (2026). AdapTools: Adaptive Tool-based Indirect Prompt Injection Attacks on Agentic LLMs. In Proceedings of the Conference on Empirical Methods in Natural Language Processing (EMNLP). Association for Computational Linguistics.

14. Zhang, F.*, Zhang, J.**, Wang, C., Sun, X., Hao, Y.**, Guan, G.*, Li, W., Huang, L., & Lim, W. Y. B. (2026). DualTAP: A dual-task adversarial protector for mobile MLLM agents. In Proceedings of the European Conference on Computer Vision (ECCV).

15. Guan, G.*, Hao**, Y., Zhang, J.**, Wu, T.**, Zhang, F.*, Chen, T., Huang, L., Leung, C. and Lim, W.Y.B., 2026. "VENOMREC: Cross-Modal Interactive Poisoning for Targeted Promotion in Multimodal LLM Recommender Systems." The Forty-third International Conference on Machine Learning (ICML).

16. Gong, Z.*, Hou, Y., Wu, F.*, Wang, C.*, Zhang, F.*, Wu, T.**, Hao, Y.**, Zhang, J.**, Duan, Y.*, Wang, T.*, Huang, F., Yuen, C. and Lim, W.Y.B., 2026. "SubspacePath Pruner: Inference-time Pruning via Probe-based Representation–Parameter Coupling." The Forty-third International Conference on Machine Learning (ICML).

17. Gong, Z.*, Wu, T.**, Zhang, J.**, Zhang, F.*, Wang, C.*, Hao, Y.**, Hou, Y.**, Foo, P.*, Zhao, Y.*, Huang, F., Yuen, C. and Lim, W.Y.B., 2026. "XDomainBench: Diagnosing Reasoning Collapse in High-Dimensional Scientific Knowledge Composition." The Forty-third International Conference on Machine Learning (ICML).

18. Mi, Y.*, Li, Y., Li, Y., Chen, H., Zhang, T., Li, Z., Song, C., Lim, W.Y.B. and Liu, S., 2026. "Q-CLIP: Unleashing the Power of Vision-Language Models for Video Quality Assessment through Unified Cross-Modal Adaptation." The Forty-third International Conference on Machine Learning (ICML).

19. Hu, X., Wang, C., Lim, W.Y.B., Gao, J. and Chen, Z., 2026. "Tracing the Dynamics of Refusal: Exploiting Latent Refusal Trajectories for Robust Jailbreak Detection." The Forty-third International Conference on Machine Learning (ICML).

20. Sun, X.*, Li, H., Zhang, J.**, Yang, Y., Liu, K., Feng, R., Tan, W.J. and Lim, W.Y.B., 2026. "MultiPriv: Benchmarking Individual-Level Privacy Reasoning in Vision-Language Models." The Forty-third International Conference on Machine Learning (ICML).

21. Wu, T.**, Hao, Y.** and Lim, W.Y.B., 2026. "An Empirical Study on the Resilience of Partial Merging to Model Clone Attacks." The Forty-third International Conference on Machine Learning (ICML).

22. Delattre, B., Wu, H., Caillon, P., Lim, W.Y.B. and Cao, Y., 2026. "Certified Robustness under Heterogeneous Perturbations via Hybrid Randomized Smoothing." The Forty-third International Conference on Machine Learning (ICML).

23. Zhang, J.**, Wang, C.*, Cao, Y., Huang, L., & Lim, W. Y. B. (2025). Disrupting Hierarchical Reasoning: Adversarial Protection for Geographic Privacy in Multimodal Reasoning Models. ICLR 2026. 

24. Zihao, Z.**, Li, S., Yan, X.*, Xiao, L., & Lim, W. Y. B. AdaCache: Adaptive Caching and Context Augmentation for Efficient LLM Serving. ICLR 2026 ##

25.  Chen, T.*, Hou, W., Wang, F., Wu, T.**, Zheng, Z., Tang, S., & Lim, W. Y. B. (2026). FedAdamom: Adaptive momentum for improved generalization in federated optimization. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Oral. 

26. Yan, B., Hao, Y.**, Liu, D., Sun, H., Qiao, P., Lim, W. Y. B., ... & Shi, C. (2025). Phantom Subgroup Poisoning: Stealth Attacks on Federated Recommender Systems. WWW 2026

27.  Wang, T.*, Duan, Y.*, Chen, H.*, Wu, T.**, & Lim, W. Y. B. (2026). M-Loss: Quantifying Model Merging Compatibility with Limited Unlabeled Data. AAAI 2026.

28. Zhang, F.*, Yan, X.*, Wu, T.**, Li, W., Chen, T., Cao, Y., Lim, W. Y. B. & Yang, Q. (2026). Oblivionis: A Lightweight Learning and Unlearning Framework for Federated Large Language Models. AAAI 2026.

29. Wang, Z.*, Yan, X.*, Wang, C., Zihao, Z., Xiao, L., & Lim, W. Y. B. SEMIROUTER: Sparse-Data Enhanced Routing for Adaptive Multi-LLM System. EACL Main Oral 2026. 

30. Zhao, Y.*, Zhang, W., Yang, T., Jiang, Y., Huang, F., & Lim, W. Y. B. STORM: A spatio-temporal factor model based on dual vector quantized variational autoencoders for financial trading. WSDM (Oral) 2026 

31.  Wu, F.*, Yan, X.*, Liu, J.*, & Lim, W. Y. B. (2025). "FedRAM: Federated Reweighting and Aggregation for Multi-Task Learning." Neural Information Processing Systems (NeurIPS), 2025. 

32.  Wang, C.*, Zhang, Z., Wang, Y., Wang, T.*, Hao, Y.**, Gao, J., Wei, T., Yang, C., Chen, Z., & Lim, W. Y. B. (2025). "AegisGuard: RL-Guided Adapter Tuning for TEE-Based Efficient & Secure On-Device Inference." Neural Information Processing Systems (NeurIPS), 2025. 

33. Zhang, J.*, Duan, Y., Niu, S., Cao, Y. and Lim, W.Y.B., 2025. Enhancing Federated Domain Adaptation with Muti-Domain Prototype-Based Federated Fine-Tuning. The Thirteenth International Conference on Learning Representations (ICLR) 

34. Zhong, Z., Bao, W., Wang, J., Zhang, S., Zhou, J., Lyu, L., & Lim, W. Y. B. (2025). Unlearning through Knowledge Overwriting: Reversible Federated Unlearning via Selective Sparse Adapter. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2025 

Journals

1. Chen, H.&, Wang, T.*, Li, Y., Lim, W.Y.B, 2026. Rethinking Model Merging: Partial Merging via Conflict Mitigation and Adaptive Weighting. IEEE Transactions on Knowledge and Data Engineering.

2. Lu, H., Pan, X., Lim, W. Y. B., & Yan, R. (2026). FedGNN-SFCP: Federated graph neural networks for ship fuel consumption prediction with heterogeneous data fusion and missing data imputation. Transportation Research Part E: Logistics and Transportation Review, 214, 105016.

3. Zhou, Z.*, Liang, Y. C., Cheng, Y., & Lim, W. Y. B. (2026). Communication-Pipelined Split Federated Learning for Foundation Model Fine-Tuning in UAV Networks. IEEE Transactions on Mobile Computing.

4. Li, W., Fan, K., Zhang, J.*, Li, H., Lim, W.Y.B. and Yang, Q., 2025. Enhancing Security and Privacy in Federated Learning using Low-Dimensional Update Representation and Proximity-Based Defense. IEEE Transactions on Knowledge and Data Engineering.(2025) 

5.  Zhong, Z., Bao, W., Wang, J., Chen, J., Lyu, L. and Lim, W.Y.B., 2025. “SacFL: Self-Adaptive Federated Continual Learning for Resource-Constrained End Devices.” IEEE Transactions on Neural Networks and Learning Systems.