Spinning Language Models: Risks of Propaganda-As-A-Service and Countermeasures von Bagdasaryan, Eugene, Shmatikov, Vitaly

Volltext

Towards Sparse Federated Analytics: Location Heatmaps under Distributed Differential Privacy with Secure Aggregation von Bagdasaryan, Eugene, Kairouz, Peter, Mellem, Stefan, Gascón, Adrià, Bonawitz, Kallista, Estrin, Deborah, Gruteser, Marco

Volltext

Backdooring Bias into Text-to-Image Models von Naseh, Ali, Roh, Jaechul, Bagdasaryan, Eugene, Houmansadr, Amir

Volltext bestellen

Mithridates: Auditing and Boosting Backdoor Resistance of Machine Learning Pipelines von Bagdasaryan, Eugene, Shmatikov, Vitaly

Volltext bestellen

Spinning Language Models: Risks of Propaganda-As-A-Service and Countermeasures von Bagdasaryan, Eugene, Shmatikov, Vitaly

Volltext

Spinning Sequence-to-Sequence Models with Meta-Backdoors von Bagdasaryan, Eugene, Shmatikov, Vitaly

Volltext bestellen

Adversarial Illusions in Multi-Modal Embeddings von Zhang, Tingwei, Jha, Rishi, Bagdasaryan, Eugene, Shmatikov, Vitaly

Volltext bestellen

Abusing Images and Sounds for Indirect Instruction Injection in Multi-Modal LLMs von Bagdasaryan, Eugene, Hsieh, Tsung-Yin, Nassi, Ben, Shmatikov, Vitaly

Volltext bestellen

Blind Backdoors in Deep Learning Models von Bagdasaryan, Eugene, Shmatikov, Vitaly

Volltext bestellen

Differential Privacy Has Disparate Impact on Model Accuracy von Bagdasaryan, Eugene, Shmatikov, Vitaly

Volltext bestellen

Salvaging Federated Learning by Local Adaptation von Yu, Tao, Bagdasaryan, Eugene, Shmatikov, Vitaly

Volltext bestellen

Training a Tokenizer for Free with Private Federated Learning von Bagdasaryan, Eugene, Song, Congzheng, van Dalen, Rogier, Seigel, Matt, Cahill, Áine

Volltext bestellen

Mithridates: Auditing and Boosting Backdoor Resistance of Machine Learning Pipelines von Bagdasaryan, Eugene, Shmatikov, Vitaly

Volltext

Policy-Based Federated Learning von Katevas, Kleomenis, Bagdasaryan, Eugene, Waterman, Jason, Safadieh, Mohamad Mounir, Birrell, Eleanor, Haddadi, Hamed, Estrin, Deborah

Volltext bestellen

Spinning Sequence-to-Sequence Models with Meta-Backdoors von Bagdasaryan, Eugene, Shmatikov, Vitaly

Volltext

How To Backdoor Federated Learning von Bagdasaryan, Eugene, Veit, Andreas, Hua, Yiqing, Estrin, Deborah, Shmatikov, Vitaly

Volltext bestellen

Blind Backdoors in Deep Learning Models von Bagdasaryan, Eugene, Shmatikov, Vitaly

Volltext

UnUnlearning: Unlearning is not sufficient for content regulation in advanced generative AI von Shumailov, Ilia, Hayes, Jamie, Triantafillou, Eleni, Ortiz-Jimenez, Guillermo, Papernot, Nicolas, Jagielski, Matthew, Yona, Itay, Howard, Heidi, Bagdasaryan, Eugene

Volltext bestellen

Operationalizing Contextual Integrity in Privacy-Conscious Assistants von Ghalebikesabi, Sahra, Bagdasaryan, Eugene, Yi, Ren, Yona, Itay, Shumailov, Ilia, Pappu, Aneesh, Shi, Chongyang, Weidinger, Laura, Stanforth, Robert, Berrada, Leonard, Kohli, Pushmeet, Huang, Po-Sen, Balle, Borja

Volltext bestellen

Backdooring Bias into Text-to-Image Models von Naseh, Ali, Roh, Jaechul, Bagdasaryan, Eugene, Houmansadr, Amir

Volltext