@inproceedings{tastan2026stochastic,title={Stochastic Self-Organization in Multi-Agent Systems},author={Tastan, Nurbek and Horv{\'a}th, Samuel and Nandakumar, Karthik},booktitle={The Fourteenth International Conference on Learning Representations},year={2026},url={https://openreview.net/forum?id=rS3Jb9AAej},}
@inproceedings{tastan2026loft,title={{Lo{FT}: Low-Rank Adaptation That Behaves Like Full Fine-Tuning}},author={Tastan, Nurbek and Laskaridis, Stefanos and Tak{\'a}{\v{c}}, Martin and Nandakumar, Karthik and Horv{\'a}th, Samuel},booktitle={The Fourteenth International Conference on Learning Representations},year={2026},url={https://openreview.net/forum?id=86P3sb1dpr},}
@inproceedings{fares2026molm,title={{MOLM}: Mixture of Lo{RA} Markers},author={Fares, Samar and Tastan, Nurbek and Hussein, Noor Hazim and Nandakumar, Karthik},booktitle={The Fourteenth International Conference on Learning Representations},year={2026},url={https://openreview.net/forum?id=1fYQOZovHR},}
@misc{tastan2026mose,title={MoSE: Mixture of Slimmable Experts for Efficient and Adaptive Language Models},author={Tastan, Nurbek and Laskaridis, Stefanos and Nandakumar, Karthik and Horvath, Samuel},year={2026},eprint={2602.06154},archiveprefix={arXiv},primaryclass={cs.LG},url={https://arxiv.org/abs/2602.06154},}
@misc{fares2025spdmark,title={SPDMark: Selective Parameter Displacement for Robust Video Watermarking},author={Fares, Samar and Tastan, Nurbek and Nandakumar, Karthik},year={2026},eprint={2512.12090},archiveprefix={arXiv},primaryclass={cs.CV},url={https://arxiv.org/abs/2512.12090},}
A Framework for Double-Blind Federated Adaptation of Foundation Models
Nurbek Tastan, and Karthik Nandakumar
In Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), ICLR 2025 Workshop on Modularity for Collaborative, Decentralized, and Continual Deep Learning (MCDC), 2025
@inproceedings{tastan2025framework,title={{A Framework for Double-Blind Federated Adaptation of Foundation Models}},author={Tastan, Nurbek and Nandakumar, Karthik},booktitle={Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), ICLR 2025 Workshop on Modularity for Collaborative, Decentralized, and Continual Deep Learning (MCDC)},year={2025},url={https://openreview.net/forum?id=stv0Fqxekz},}
@article{tastan2025cycle,title={{{CYC}le: Choosing Your Collaborators Wisely to Enhance Collaborative Fairness in Decentralized Learning}},author={Tastan, Nurbek and Horv{\'a}th, Samuel and Nandakumar, Karthik},journal={Transactions on Machine Learning Research},issn={2835-8856},year={2025},url={https://openreview.net/forum?id=ygqNiLQqfH},note={},}
@inproceedings{tastan2025aequa,title={{Aequa: Fair Model Rewards in Collaborative Learning via Slimmable Networks}},author={Tastan, Nurbek and Horv\'{a}th, Samuel and Nandakumar, Karthik},year={2025},url={https://openreview.net/forum?id=Tw81RElDpe},booktitle={Proceedings of the Forty-second International Conference on Machine Learning (ICML)},series={Proceedings of Machine Learning Research},}
Statistical data heterogeneity is a significant barrier to convergence in federated learning (FL). While prior work has advanced heterogeneous FL through better optimization objectives, these methods fall short when there is *extreme* data heterogeneity among collaborating participants. We hypothesize that convergence under extreme data heterogeneity is primarily hindered due to the aggregation of conflicting updates from the participants in the initial collaboration rounds. To overcome this problem, we propose a warmup phase where each participant learns a personalized mask and updates only a subnetwork of the full model. This *personalized warmup* allows the participants to focus initially on learning specific *subnetworks* tailored to the heterogeneity of their data. After the warmup phase, the participants revert to standard federated optimization, where all parameters are communicated. We empirically demonstrate that the proposed personalized warmup via subnetworks (*FedPeWS*) approach improves accuracy and convergence speed over standard federated optimization methods. The code can be found at https://github.com/tnurbek/fedpews.
@inproceedings{tastan2025fedpews,title={{FedPeWS: Personalized Warmup via Subnetworks for Enhanced Heterogeneous Federated Learning}},author={Tastan, Nurbek and Horv\'{a}th, Samuel and Tak\'{a}\v{c}, Martin and Nandakumar, Karthik},booktitle={Conference on Parsimony and Learning},pages={462--483},year={2025},editor={Chen, Beidi and Liu, Shijia and Pilanci, Mert and Su, Weijie and Sulam, Jeremias and Wang, Yuxiang and Zhu, Zhihui},volume={280},series={Proceedings of Machine Learning Research},month={24--27 Mar},publisher={PMLR},url={https://proceedings.mlr.press/v280/tastan25a.html},}
Federated learning (FL) has emerged as a pivotal approach in machine learning, enabling multiple participants to collaboratively train a global model without sharing raw data. While FL finds applications in various domains such as healthcare and finance, it is challenging to ensure global model convergence when participants do not contribute equally and/or honestly. To overcome this challenge, principled mechanisms are required to evaluate the contributions made by individual participants in the FL setting. Existing solutions for contribution assessment rely on general accuracy evaluation, often failing to capture nuanced dynamics and class-specific influences. This paper proposes a novel contribution assessment method called ShapFed for fine-grained evaluation of participant contributions in FL. Our approach uses Shapley values from cooperative game theory to provide a granular understanding of class-specific influences. Based on ShapFed, we introduce a weighted aggregation method called ShapFed-WA, which outperforms conventional federated averaging, especially in class-imbalanced scenarios. Personalizing participant updates based on their contributions further enhances collaborative fairness by delivering differentiated models commensurate with the participant contributions. Experiments on CIFAR-10, Chest X-Ray, and Fed-ISIC2019 datasets demonstrate the effectiveness of our approach in improving utility, efficiency, and fairness in FL systems.
@inproceedings{tastan2024redefining,author={Tastan, Nurbek and Fares, Samar and Aremu, Toluwani and Horvath, Samuel and Nandakumar, Karthik},title={Redefining Contributions: Shapley-Driven Federated Learning},booktitle={International Joint Conference on Artificial Intelligence (IJCAI)},year={2024},month=aug,}
@inproceedings{al2024collaborative,title={Collaborative Learning of Anomalies with Privacy (CLAP) for Unsupervised Video Anomaly Detection: A New Baseline},author={Al-lahham, Anas and Zaheer, Muhammad Zaigham and Tastan, Nurbek and Nandakumar, Karthik},booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},year={2024},month=jun,pages={12416-12425},}
@inproceedings{al2024coarse,title={A Coarse-to-Fine Pseudo-Labeling (C2FPL) Framework for Unsupervised Video Anomaly Detection},author={Al-lahham, Anas and Tastan, Nurbek and Zaheer, Muhammad Zaigham and Nandakumar, Karthik},booktitle={Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision},pages={6793--6802},year={2024},month=jan,}
@inproceedings{tastan2023capride,author={Tastan, Nurbek and Nandakumar, Karthik},title={CaPriDe Learning: Confidential and Private Decentralized Learning Based on Encryption-Friendly Distillation Loss},booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},month=jun,year={2023},pages={8084-8092},}
