Edoardo Debenedetti

Researcher, Meta

PhD student, Computer Science, ETH Zurich

Researcher, AI Sequrity Company

Joined

October 2019

Names

Edoardo Debenedetti (Preferred)

Emails

****@epfl.ch (Confirmed)

****@gmail.com (Confirmed)

****@inf.ethz.ch (Confirmed)

****@google.com (Confirmed)

****@meta.com (Confirmed)

****@sequrity.ai (Confirmed)

Personal Links

Career & Education History

Researcher

Meta (meta.com)

2026 – Present

PhD student

Computer Science, ETH Zurich (ethz.ch)

2022 – Present

Researcher

AI Sequrity Company (sequrity.ai)

2026 – 2026

Intern

Meta (meta.com)

2025 – 2025

Intern

Google (google.com)

2024 – 2025

MS student

Computer and Communication Sciences, EPFL (epfl.ch)

2019 – 2022

Undergrad student

DAUIN, Politecnico di Torino (polito.it)

2016 – 2019

Advisors, Relations & Conflicts

PhD Advisor

Florian Tramèr

2022 – Present

Expertise

machine learning

adversarial examples

evasion attacks

2021 – Present

Publications

Defeating Prompt Injections by Design
Edoardo Debenedetti, Ilia Shumailov, Tianqi Fan, Jamie Hayes, Nicholas Carlini, Daniel Fabian, Christoph Kern, Chongyang Shi, Andreas Terzis, Florian Tramèr
- CoRR 2025
- Readers: Everyone
AutoAdvExBench: Benchmarking Autonomous Exploitation of Adversarial Example Defenses
Nicholas Carlini, Edoardo Debenedetti, Javier Rando, Milad Nasr, Florian Tramèr
- ICML 2025 oral
- Readers: Everyone
Exploring Memorization and Copyright Violation in Frontier LLMs: A Study of the New York Times v. OpenAI 2023 Lawsuit
Joshua Freeman, Chloe Rippe, Edoardo Debenedetti, Maksym Andriushchenko
- SafeGenAi Poster
- Readers: Everyone
AutoAdvExBench: Benchmarking Autonomous Exploitation of Adversarial Example Defenses
Nicholas Carlini, Edoardo Debenedetti, Javier Rando, Milad Nasr, Florian Tramèr
- Submitted to ICLR 2025
- Readers: Everyone
Measuring Non-Adversarial Reproduction of Training Data in Large Language Models
Michael Aerni, Javier Rando, Edoardo Debenedetti, Nicholas Carlini, Daphne Ippolito, Florian Tramèr
- ICLR 2025 Poster
- Readers: Everyone
Adversarial Search Engine Optimization for Large Language Models
Fredrik Nestaas, Edoardo Debenedetti, Florian Tramèr
- ICLR 2025 Poster
- Readers: Everyone
JailbreakBench: An Open Robustness Benchmark for Jailbreaking Large Language Models
Patrick Chao, Edoardo Debenedetti, Alexander Robey, Maksym Andriushchenko, Francesco Croce, Vikash Sehwag, Edgar Dobriban, Nicolas Flammarion, George J. Pappas, Florian Tramèr, Hamed Hassani, Eric Wong
- NextGenAISafety 2024 Poster
- Readers: Everyone
AgentDojo: A Dynamic Environment to Evaluate Prompt Injection Attacks and Defenses for LLM Agents
Edoardo Debenedetti, Jie Zhang, Mislav Balunovic, Luca Beurer-Kellner, Marc Fischer, Florian Tramèr
- NeurIPS 2024 Track Datasets and Benchmarks Poster
- Readers: Everyone
JailbreakBench: An Open Robustness Benchmark for Jailbreaking Large Language Models
Patrick Chao, Edoardo Debenedetti, Alexander Robey, Maksym Andriushchenko, Francesco Croce, Vikash Sehwag, Edgar Dobriban, Nicolas Flammarion, George J. Pappas, Florian Tramèr, Hamed Hassani, Eric Wong
- NeurIPS 2024 Track Datasets and Benchmarks Poster
- Readers: Everyone
Dataset and Lessons Learned from the 2024 SaTML LLM Capture-the-Flag Competition
Edoardo Debenedetti, Javier Rando, Daniel Paleka, Silaghi Fineas Florin, Dragos Albastroiu, Niv Cohen, Yuval Lemberg, Reshmi Ghosh, Rui Wen, Ahmed Salem, Giovanni Cherubin, Santiago Zanella-Beguelin, Robin Schmid, Victor Klemm, Takahiro Miki, Chenhao Li, Stefan Kraft, Mario Fritz, Florian Tramèr, Sahar Abdelnabi et al. (1 additional authors not shown)
- NeurIPS 2024 Track Datasets and Benchmarks Spotlight
- Readers: Everyone

View all 27 publications

Co-Authors

View all 81 co-authors