Below is a selected list of my scholarly outputs. For a complete and always-up-to-date record please refer to my Google Scholar profile.
Journal & Conference Papers
• Learning to Collaborate: An Orchestrated-Decentralized Framework for Peer-to-Peer LLM Federation (with Eleonore Vissol-Gaudin, Andikan Otung, Motoyoshi Sekiya). In Proceedings of the 2026 AAAI Conference on Artificial Intelligence (AAAI).
• TFDP: Token-Efficient Disparity Audits for Autoregressive LLMs via Single-Token Masked Evaluation (with Ramya Srinivasan, Roman Vainshtein, Hisashi Kojima). In Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing (EMNLP), pp. 24609-24626, Suzhou, China. ACL Anthology | PDF | DOI
• Evaluating the Cybersecurity Risk of Real-world, Machine Learning Production Systems. ACM Computing Surveys 55 (9), 1-36, 2023. Publisher | arXiv
• A Framework for Evaluating the Cybersecurity Risk of Real-world, Machine Learning Production Systems. arXiv preprint arXiv:2107.01806, 2021. PDF
• Dodging Attack Using Carefully Crafted Natural Makeup. arXiv preprint arXiv:2109.06467, 2021. PDF
• Toward Practical Adversarial Attacks on Face Verification Systems. In Proc. BIOSIG 2021, pp. 1-6, IEEE. PDF
• On Brightness-Agnostic Adversarial Examples Against Face Recognition Systems. In Proc. BIOSIG 2021, pp. 1-5, IEEE. IEEE Xplore
• Powerful Physical Adversarial Examples Against Practical Face Recognition Systems. In Proc. WACV 2022, pp. 301-310. OpenAccess
• Simultaneous Adversarial Attacks on Multiple Face Recognition System Components. arXiv preprint arXiv:2304.05048, 2023. PDF
• Anomaly Detection Using Capsule Networks for High-dimensional Datasets. arXiv preprint arXiv:2112.13514, 2021. PDF
Open Datasets & Benchmarks
• Agentic RAG Red Teaming Dataset (with Vikas Pahuja, Aishvariya Priya Rathina Sabapathy). Fujitsu Research, 2025. 36K+ successful adversarial RAG scenarios spanning retrieval poisoning, direct queries, agent orchestration, and OCR-mediated attacks for safety evaluation. Hugging Face | DOI
• PDD-Extended Bench (Proverbs Disparity Dataset) (with Ramya Srinivasan, Roman Vainshtein, Hisashi Kojima). EMNLP 2025 companion dataset. Minimally contrastive, single-token masked proverbs across climate misinformation and gender equality topics for TFDP disparity audits. Hugging Face | DOI
Patents
• Facial Recognition Adversarial Patch Adjustment. US Patent, 2024. USPTO
• Adversarial Image Generation Apparatus, Control Method, and Computer-readable Storage Medium. WO 2023 286251 A1, 2023. PatentScope