Saeed Mahloujifar

100 papers A* 28A 1B 3C 1Misc 3Journal 60Unranked 4

Year	Rank	Type	Title / Venue / Authors
2026	J	`jnl`	Learning to Reason in 13 Parameters. CoRR John X. Morris, Niloofar Mireshghallah, Mark Ibrahim, Saeed Mahloujifar
2025	A*	`conf`	Auditing $f$-differential privacy in one run. ICML Saeed Mahloujifar, Luca Melis, Kamalika Chaudhuri
2025	J	`jnl`	CIMemories: A Compositional Benchmark for Contextual Integrity of Persistent Memory in LLMs. CoRR Niloofar Mireshghallah, Neal Mangaokar, Narine Kokhlikyan, Arman Zharmagambetov, Manzil Zaheer, Saeed Mahloujifar, Kamalika Chaudhuri
2025	J	`jnl`	Detecting Benchmark Contamination Through Watermarking. CoRR Tom Sander, Pierre Fernandez, Saeed Mahloujifar, Alain Durmus, Chuan Guo
2025	J	`jnl`	How much do language models memorize? CoRR John X. Morris, Chawin Sitawarin, Chuan Guo, Narine Kokhlikyan, G. Edward Suh, Alexander M. Rush, Kamalika Chaudhuri, Saeed Mahloujifar
2025	A*	`conf`	Machine Learning with Privacy for Protected Attributes. SP Saeed Mahloujifar, Chuan Guo, G. Edward Suh, Kamalika Chaudhuri
2025	J	`jnl`	Machine Learning with Privacy for Protected Attributes. CoRR Saeed Mahloujifar, Chuan Guo, G. Edward Suh, Kamalika Chaudhuri
2025	J	`jnl`	Observational Auditing of Label Privacy. CoRR Iden Kalemaj, Luca Melis, Maxime Boucher, Ilya Mironov, Saeed Mahloujifar
2025	J	`jnl`	Privacy Blur: Quantifying Privacy and Utility for Image Data Release. CoRR Saeed Mahloujifar, Narine Kokhlikyan, Chuan Guo, Kamalika Chaudhuri
2025	J	`jnl`	Private Fine-tuning of Large Language Models with Zeroth-order Optimization. Trans. Mach. Learn. Res. Xinyu Tang, Ashwinee Panda, Milad Nasr, Saeed Mahloujifar, Prateek Mittal
2025	A*	`conf`	SecAlign: Defending Against Prompt Injection with Preference Optimization. CCS Sizhe Chen, Arman Zharmagambetov, Saeed Mahloujifar, Kamalika Chaudhuri, David A. Wagner, Chuan Guo
2025	J	`jnl`	Unlocking Visual Secrets: Inverting Features with Diffusion Priors for Image Reconstruction. Trans. Mach. Learn. Res. Sai Qian Zhang, Ziyun Li, Chuan Guo, Saeed Mahloujifar, Deeksha Dangwal, G. Edward Suh, Barbara De Salvo, Chiao Liu
2025	J	`jnl`	Z0-Inf: Zeroth Order Approximation for Data Influence. CoRR Narine Kokhlikyan, Kamalika Chaudhuri, Saeed Mahloujifar
2024	A*	`conf`	A New Linear Scaling Rule for Private Adaptive Hyperparameter Optimization. ICML Ashwinee Panda, Xinyu Tang, Saeed Mahloujifar, Vikash Sehwag, Prateek Mittal
2024	J	`jnl`	Aligning LLMs to Be Robust Against Prompt Injection. CoRR Sizhe Chen, Arman Zharmagambetov, Saeed Mahloujifar, Kamalika Chaudhuri, Chuan Guo
2024	J	`jnl`	Auditing CoRR Saeed Mahloujifar, Luca Melis, Kamalika Chaudhuri
2024	J	`jnl`	Guarantees of confidentiality via Hammersley-Chapman-Robbins bounds. CoRR Kamalika Chaudhuri, Chuan Guo, Laurens van der Maaten, Saeed Mahloujifar, Mark Tygert
2024	J	`jnl`	Guarantees of confidentiality via Hammersley-Chapman-Robbins bounds. Trans. Mach. Learn. Res. Kamalika Chaudhuri, Chuan Guo, Laurens van der Maaten, Saeed Mahloujifar, Mark Tygert
2024	Misc	`conf`	Horus: Granular In-Network Task Scheduler for Cloud Datacenters. NSDI Parham Yassini, Khaled Diab, Saeed Mahloujifar, Mohamed Hefeeda
2024	J	`jnl`	Privacy Amplification for the Gaussian Mechanism via Bounded Support. CoRR Shengyuan Hu, Saeed Mahloujifar, Virginia Smith, Kamalika Chaudhuri, Chuan Guo
2024	J	`jnl`	Private Fine-tuning of Large Language Models with Zeroth-order Optimization. CoRR Xinyu Tang, Ashwinee Panda, Milad Nasr, Saeed Mahloujifar, Prateek Mittal
2024	J	`jnl`	Publicly-Detectable Watermarking for Language Models. IACR Commun. Cryptol. Jaiden Fairoze, Sanjam Garg, Somesh Jha, Saeed Mahloujifar, Mohammad Mahmoody, Mingyuan Wang
2024	J	`jnl`	Unlocking Visual Secrets: Inverting Features with Diffusion Priors for Image Reconstruction. CoRR Sai Qian Zhang, Ziyun Li, Chuan Guo, Saeed Mahloujifar, Deeksha Dangwal, G. Edward Suh, Barbara De Salvo, Chiao Liu
2023	J	`jnl`	A Randomized Approach for Tight Privacy Accounting. CoRR Jiachen T. Wang, Saeed Mahloujifar, Tong Wu, Ruoxi Jia, Prateek Mittal
2023	A*	`conf`	A Randomized Approach to Tight Privacy Accounting. NeurIPS Jiachen T. Wang, Saeed Mahloujifar, Tong Wu, Ruoxi Jia, Prateek Mittal
2023	J	`jnl`	Bounding Training Data Reconstruction in DP-SGD. CoRR Jamie Hayes, Saeed Mahloujifar, Borja Balle
2023	A*	`conf`	Bounding training data reconstruction in DP-SGD. NeurIPS Jamie Hayes, Borja Balle, Saeed Mahloujifar
2023	A*	`conf`	Effectively Using Public Data in Privacy Preserving Machine Learning. ICML Milad Nasr, Saeed Mahloujifar, Xinyu Tang, Prateek Mittal, Amir Houmansadr
2023	A*	`conf`	Experimenting with Zero-Knowledge Proofs of Training. CCS Sanjam Garg, Aarushi Goel, Somesh Jha, Saeed Mahloujifar, Mohammad Mahmoody, Guru-Vamsi Policharla, Mingyuan Wang
2023	J	`jnl`	Experimenting with Zero-Knowledge Proofs of Training. IACR Cryptol. ePrint Arch. Sanjam Garg, Aarushi Goel, Somesh Jha, Saeed Mahloujifar, Mohammad Mahmoody, Guru-Vamsi Policharla, Mingyuan Wang
2023	A*	`conf`	MultiRobustBench: Benchmarking Robustness Against Multiple Attacks. ICML Sihui Dai, Saeed Mahloujifar, Chong Xiang, Vikash Sehwag, Pin-Yu Chen, Prateek Mittal
2023	J	`jnl`	MultiRobustBench: Benchmarking Robustness Against Multiple Attacks. CoRR Sihui Dai, Saeed Mahloujifar, Chong Xiang, Vikash Sehwag, Pin-Yu Chen, Prateek Mittal
2023	A*	`conf`	ObjectSeeker: Certifiably Robust Object Detection against Patch Hiding Attacks via Patch-agnostic Masking. SP Chong Xiang, Alexander Valtchanov, Saeed Mahloujifar, Prateek Mittal
2023	J	`jnl`	Publicly Detectable Watermarking for Language Models. CoRR Jaiden Fairoze, Sanjam Garg, Somesh Jha, Saeed Mahloujifar, Mohammad Mahmoody, Mingyuan Wang
2023	J	`jnl`	Publicly Detectable Watermarking for Language Models. IACR Cryptol. ePrint Arch. Jaiden Fairoze, Sanjam Garg, Somesh Jha, Saeed Mahloujifar, Mohammad Mahmoody, Mingyuan Wang
2023	A*	`conf`	Revisiting the Assumption of Latent Separability for Backdoor Defenses. ICLR Xiangyu Qi, Tinghao Xie, Yiming Li, Saeed Mahloujifar, Prateek Mittal
2023	A*	`conf`	Towards A Proactive ML Approach for Detecting Backdoor Poison Samples. USENIX Security Symposium Xiangyu Qi, Tinghao Xie, Jiachen T. Wang, Tong Wu, Saeed Mahloujifar, Prateek Mittal
2023	A*	`conf`	Uncovering Adversarial Risks of Test-Time Adaptation. ICML Tong Wu, Feiran Jia, Xiangyu Qi, Jiachen T. Wang, Vikash Sehwag, Saeed Mahloujifar, Prateek Mittal
2023	J	`jnl`	Uncovering Adversarial Risks of Test-Time Adaptation. CoRR Tong Wu, Feiran Jia, Xiangyu Qi, Jiachen T. Wang, Vikash Sehwag, Saeed Mahloujifar, Prateek Mittal
2022	J	`jnl`	Circumventing Backdoor Defenses That Are Based on Latent Separability. CoRR Xiangyu Qi, Tinghao Xie, Saeed Mahloujifar, Prateek Mittal
2022	J	`jnl`	DP-RAFT: A Differentially Private Recipe for Accelerated Fine-Tuning. CoRR Ashwinee Panda, Xinyu Tang, Vikash Sehwag, Saeed Mahloujifar, Prateek Mittal
2022	J	`jnl`	Fight Poison with Poison: Detecting Backdoor Poison Samples via Decoupling Benign Correlations. CoRR Xiangyu Qi, Tinghao Xie, Saeed Mahloujifar, Prateek Mittal
2022	A*	`conf`	Formulating Robustness Against Unforeseen Attacks. NeurIPS Sihui Dai, Saeed Mahloujifar, Prateek Mittal
2022	J	`jnl`	Formulating Robustness Against Unforeseen Attacks. CoRR Sihui Dai, Saeed Mahloujifar, Prateek Mittal
2022	—	`conf`	Just Rotate it: Deploying Backdoor Attacks via Rotation Transformation. AISec@CCS Tong Wu, Tianhao Wang, Vikash Sehwag, Saeed Mahloujifar, Prateek Mittal
2022	J	`jnl`	Just Rotate it: Deploying Backdoor Attacks via Rotation Transformation. CoRR Tong Wu, Tianhao Wang, Vikash Sehwag, Saeed Mahloujifar, Prateek Mittal
2022	J	`jnl`	Machine Learning with Differentially Private Labels: Mechanisms and Frameworks. Proc. Priv. Enhancing Technol. Xinyu Tang, Milad Nasr, Saeed Mahloujifar, Virat Shejwalkar, Liwei Song, Amir Houmansadr, Prateek Mittal
2022	A*	`conf`	Mitigating Membership Inference Attacks by Self-Distillation Through a Novel Ensemble Architecture. USENIX Security Symposium Xinyu Tang, Saeed Mahloujifar, Liwei Song, Virat Shejwalkar, Milad Nasr, Amir Houmansadr, Prateek Mittal
2022	J	`jnl`	ObjectSeeker: Certifiably Robust Object Detection against Patch Hiding Attacks via Patch-agnostic Masking. CoRR Chong Xiang, Alexander Valtchanov, Saeed Mahloujifar, Prateek Mittal
2022	J	`jnl`	Optimal Membership Inference Bounds for Adaptive Composition of Sampled Gaussian Mechanisms. CoRR Saeed Mahloujifar, Alexandre Sablayrolles, Graham Cormode, Somesh Jha
2022	A*	`conf`	Overparameterization from Computational Constraints. NeurIPS Sanjam Garg, Somesh Jha, Saeed Mahloujifar, Mohammad Mahmoody, Mingyuan Wang
2022	J	`jnl`	Overparameterized (robust) models from computational constraints. CoRR Sanjam Garg, Somesh Jha, Saeed Mahloujifar, Mohammad Mahmoody, Mingyuan Wang
2022	—	`conf`	Parameterizing Activation Functions for Adversarial Robustness. SP (Workshops) Sihui Dai, Saeed Mahloujifar, Prateek Mittal
2022	A*	`conf`	PatchCleanser: Certifiably Robust Defense against Adversarial Patches for Any Image Classifier. USENIX Security Symposium Chong Xiang, Saeed Mahloujifar, Prateek Mittal
2022	A*	`conf`	Property Inference from Poisoning. SP Saeed Mahloujifar, Esha Ghosh, Melissa Chase
2022	A*	`conf`	Renyi Differential Privacy of Propose-Test-Release and Applications to Private and Robust Machine Learning. NeurIPS Jiachen T. Wang, Saeed Mahloujifar, Shouda Wang, Ruoxi Jia, Prateek Mittal
2022	J	`jnl`	Renyi Differential Privacy of Propose-Test-Release and Applications to Private and Robust Machine Learning. CoRR Jiachen T. Wang, Saeed Mahloujifar, Shouda Wang, Ruoxi Jia, Prateek Mittal
2022	A*	`conf`	Robust Learning Meets Generative Models: Can Proxy Distributions Improve Adversarial Robustness? ICLR Vikash Sehwag, Saeed Mahloujifar, Tinashe Handina, Sihui Dai, Chong Xiang, Mung Chiang, Prateek Mittal
2022	A	`conf`	SparseFed: Mitigating Model Poisoning Attacks in Federated Learning with Sparsification. AISTATS Ashwinee Panda, Saeed Mahloujifar, Arjun Nitin Bhagoji, Supriyo Chakraborty, Prateek Mittal
2021	A*	`conf`	A Separation Result Between Data-oblivious and Data-aware Poisoning Attacks. NeurIPS Samuel Deng, Sanjam Garg, Somesh Jha, Saeed Mahloujifar, Mohammad Mahmoody, Abhradeep Guha Thakurta
2021	J	`jnl`	Improving Adversarial Robustness Using Proxy Distributions. CoRR Vikash Sehwag, Saeed Mahloujifar, Tinashe Handina, Sihui Dai, Chong Xiang, Mung Chiang, Prateek Mittal
2021	A*	`conf`	Is Private Learning Possible with Instance Encoding? SP Nicholas Carlini, Samuel Deng, Sanjam Garg, Somesh Jha, Saeed Mahloujifar, Mohammad Mahmoody, Abhradeep Thakurta, Florian Tramèr
2021	J	`jnl`	Membership Inference on Word Embedding and Beyond. CoRR Saeed Mahloujifar, Huseyin A. Inan, Melissa Chase, Esha Ghosh, Marcello Hasegawa
2021	J	`jnl`	Mitigating Membership Inference Attacks by Self-Distillation Through a Novel Ensemble Architecture. CoRR Xinyu Tang, Saeed Mahloujifar, Liwei Song, Virat Shejwalkar, Milad Nasr, Amir Houmansadr, Prateek Mittal
2021	A*	`conf`	Model-Targeted Poisoning Attacks with Provable Convergence. ICML Fnu Suya, Saeed Mahloujifar, Anshuman Suri, David Evans, Yuan Tian
2021	J	`jnl`	NeuraCrypt is not private. CoRR Nicholas Carlini, Sanjam Garg, Somesh Jha, Saeed Mahloujifar, Mohammad Mahmoody, Florian Tramèr
2021	J	`jnl`	Parameterizing Activation Functions for Adversarial Robustness. CoRR Sihui Dai, Saeed Mahloujifar, Prateek Mittal
2021	J	`jnl`	PatchCleanser: Certifiably Robust Defense against Adversarial Patches for Any Image Classifier. CoRR Chong Xiang, Saeed Mahloujifar, Prateek Mittal
2021	—	`conf`	Polynomial-Time Targeted Attacks on Coin Tossing for Any Number of Corruptions. TCC (2) Omid Etesami, Ji Gao, Saeed Mahloujifar, Mohammad Mahmoody
2021	J	`jnl`	Polynomial-time targeted attacks on coin tossing for any number of corruptions. IACR Cryptol. ePrint Arch. Omid Etesami, Ji Gao, Saeed Mahloujifar, Mohammad Mahmoody
2021	J	`jnl`	Property Inference From Poisoning. CoRR Melissa Chase, Esha Ghosh, Saeed Mahloujifar
2021	J	`jnl`	Property Inference from Poisoning. IACR Cryptol. ePrint Arch. Melissa Chase, Esha Ghosh, Saeed Mahloujifar
2021	J	`jnl`	SparseFed: Mitigating Model Poisoning Attacks in Federated Learning with Sparsification. CoRR Ashwinee Panda, Saeed Mahloujifar, Arjun Nitin Bhagoji, Supriyo Chakraborty, Prateek Mittal
2020	B	`conf`	Adversarially Robust Learning Could Leverage Computational Hardness. ALT Sanjam Garg, Somesh Jha, Saeed Mahloujifar, Mohammad Mahmoody
2020	J	`jnl`	An Attack on InstaHide: Is Private Learning Possible with Instance Encoding? CoRR Nicholas Carlini, Samuel Deng, Sanjam Garg, Somesh Jha, Saeed Mahloujifar, Mohammad Mahmoody, Shuang Song, Abhradeep Thakurta, Florian Tramèr
2020	A*	`conf`	Computational Concentration of Measure: Optimal Bounds, Reductions, and More. SODA Omid Etesami, Saeed Mahloujifar, Mohammad Mahmoody
2020	J	`jnl`	Learning under p-tampering poisoning attacks. Ann. Math. Artif. Intell. Saeed Mahloujifar, Dimitrios I. Diochnos, Mohammad Mahmoody
2020	C	`conf`	Lower Bounds for Adversarially Robust PAC Learning under Evasion and Hybrid Attacks. ICMLA Dimitrios I. Diochnos, Saeed Mahloujifar, Mohammad Mahmoody
2020	Misc	`conf`	Lower Bounds for Adversarially Robust PAC Learning. ISAIM Dimitrios I. Diochnos, Saeed Mahloujifar, Mohammad Mahmoody
2020	J	`jnl`	Model-Targeted Poisoning Attacks: Provable Convergence and Certified Bounds. CoRR Fnu Suya, Saeed Mahloujifar, David Evans, Yuan Tian
2020	J	`jnl`	Obliviousness Makes Poisoning Adversaries Weaker. CoRR Sanjam Garg, Somesh Jha, Saeed Mahloujifar, Mohammad Mahmoody, Abhradeep Thakurta
2019	J	`jnl`	Adversarially Robust Learning Could Leverage Computational Hardness. CoRR Sanjam Garg, Somesh Jha, Saeed Mahloujifar, Mohammad Mahmoody
2019	B	`conf`	Can Adversarially Robust Learning LeverageComputational Hardness? ALT Saeed Mahloujifar, Mohammad Mahmoody
2019	J	`jnl`	Computational Concentration of Measure: Optimal Bounds, Reductions, and More. CoRR Omid Etesami, Saeed Mahloujifar, Mohammad Mahmoody
2019	A*	`conf`	Data Poisoning Attacks in Multi-Party Learning. ICML Saeed Mahloujifar, Mohammad Mahmoody, Ameer Mohammed
2019	A*	`conf`	Empirically Measuring Concentration: Fundamental Limits on Intrinsic Robustness. NeurIPS Saeed Mahloujifar, Xiao Zhang, Mohammad Mahmoody, David Evans
2019	J	`jnl`	Empirically Measuring Concentration: Fundamental Limits on Intrinsic Robustness. CoRR Saeed Mahloujifar, Xiao Zhang, Mohammad Mahmoody, David Evans
2019	J	`jnl`	Lower Bounds for Adversarially Robust PAC Learning. CoRR Dimitrios I. Diochnos, Saeed Mahloujifar, Mohammad Mahmoody
2019	A*	`conf`	The Curse of Concentration in Robust Learning: Evasion and Poisoning Attacks from Concentration of Measure. AAAI Saeed Mahloujifar, Dimitrios I. Diochnos, Mohammad Mahmoody
2018	A*	`conf`	Adversarial Risk and Robustness: General Definitions and Implications for the Uniform Distribution. NeurIPS Dimitrios I. Diochnos, Saeed Mahloujifar, Mohammad Mahmoody
2018	J	`jnl`	Adversarial Risk and Robustness: General Definitions and Implications for the Uniform Distribution. CoRR Dimitrios I. Diochnos, Saeed Mahloujifar, Mohammad Mahmoody
2018	J	`jnl`	Can Adversarially Robust Learning Leverage Computational Hardness? CoRR Saeed Mahloujifar, Mohammad Mahmoody
2018	B	`conf`	Learning under $p$-Tampering Attacks. ALT Saeed Mahloujifar, Dimitrios I. Diochnos, Mohammad Mahmoody
2018	Misc	`conf`	Learning under p-Tampering Attacks. ISAIM Saeed Mahloujifar, Dimitrios I. Diochnos, Mohammad Mahmoody
2018	J	`jnl`	Multi-party Poisoning through Generalized p-Tampering. CoRR Saeed Mahloujifar, Mohammad Mahmoody, Ameer Mohammed
2018	J	`jnl`	Multi-party Poisoning through Generalized p-Tampering. IACR Cryptol. ePrint Arch. Saeed Mahloujifar, Mohammad Mahmoody, Ameer Mohammed
2018	J	`jnl`	The Curse of Concentration in Robust Learning: Evasion and Poisoning Attacks from Concentration of Measure. CoRR Saeed Mahloujifar, Dimitrios I. Diochnos, Mohammad Mahmoody
2017	—	`conf`	Blockwise p-Tampering Attacks on Cryptographic Primitives, Extractors, and Learners. TCC (2) Saeed Mahloujifar, Mohammad Mahmoody
2017	J	`jnl`	Blockwise p-Tampering Attacks on Cryptographic Primitives, Extractors, and Learners. IACR Cryptol. ePrint Arch. Saeed Mahloujifar, Mohammad Mahmoody
2017	J	`jnl`	Learning under p-Tampering Attacks. CoRR Saeed Mahloujifar, Dimitrios I. Diochnos, Mohammad Mahmoody