FT

Florian Tramèr

0000-0001-8703-8762

cs.LGcs.CRcs.AIcs.CLstat.MLcs.CVcs.CYbeaconcs.IRData Anonymization

On Valency

published · living versions

Slalom: Fast, Verifiable and Private Execution of Neural Networks in Trusted Hardware

with Dan Boneh

1 version

W_3f9m5q78v1 · current

Can Open Be Too Open? On the Security and Trustworthiness of Fully Transparent LLMs

W_t69j7389v1

Preprints & journals

106 papers in the corpus · 2015–2026

Render Before Reading: Visual Rendering as a Prompt Injection Defense

Jie Zhang, Andrei Baroian, Jan N. van Rijn et al.

0 citations

2609.36121v12026arXiv

Calibrating One-Round Membership Inference with Neighbors

Francesco Rita, Jie Zhang, Florian Tramer

0 citations

2609.36331v12026arXiv

CryptanalysisBench: Can LLMs do Cryptanalysis?

Lukas Fluri, Avital Shafran, Nicholas Carlini et al.

0 citations

2607.18538v22026arXiv

Untrusted Content Masking for Web Agents with Security Guarantees

Kristina Nikoli'c, Egor Zverev, Javier Rando et al.

1 citation

2607.05277v12026arXiv

What Does It Mean to Break a Distillation Defense?

Lena Libon, Pura Peetathawatchai, Michael Aerni et al.

0 citations

2606.25059v22026arXiv

Assessing Automated Prompt Injection Attacks in Agentic Environments

David Hofer, Edoardo Debenedetti, Florian Tramer

0 citations

2606.10525v12026arXiv

Position: Anthropomorphic Misalignment Research Needs Stronger Evidence

Vansh Gupta, Peter Nutter, Samuel Stante et al.

0 citations

2606.07612v12026arXiv

Decomposing and Measuring Evaluation Awareness

Changling Li, Terry Jingchen Zhang, Jie Zhang et al.

0 citations

2605.23055v32026arXiv

Laundering AI Authority with Adversarial Examples

Jie Zhang, Pura Peetathawatchai, Florian Tramer et al.

0 citations

2605.04261v12026arXiv

Trojan Hippo Bench: A Dynamic Benchmark for Persistent Memory Attacks and Defenses in LLM Agents

Debeshee Das, Julien Piet, Darya Kaviani et al.

0 citations

2605.01970v42026arXiv

Large-scale online deanonymization with LLMs

Simon Lermen, Daniel Paleka, Joshua Swanson et al.

1 citation

2602.16800v2202635th USENIX Security Symposium (USENIX Security 26), pp. 1947-1966 (2026)

Learning to Inject: Automated Prompt Injection via Reinforcement Learning

Xin Chen, Jie Zhang, Florian Tramer

0 citations

2602.05746v22026arXiv

Reliable and Responsible Foundation Models: A Comprehensive Survey

Xinyu Yang, Junlin Han, Rishi Bommasani et al.

0 citations

2602.08145v12026arXiv

Rethinking Benchmarks for Differentially Private Image Classification

Sabrina Mokhtari, Sara Kodeiri, Shubhankar Mohapatra et al.

0 citations

2601.17189v22026IEEE Technical Committee on Data Engineering 2025

CaMeLs Can Use Computers Too: System-level Security for Computer Use Agents

Hanna Foerster, Tom Blanchard, Kristina Nikoli'c et al.

0 citations

2601.09923v32026arXiv

Representations of Text and Images Align From Layer One

Evzen Wybitul, Javier Rando, Florian Tramer et al.

0 citations

2601.08017v12026arXiv

Modal Aphasia: Can Unified Multimodal Models Describe Images From Memory?

Michael Aerni, Joshua Swanson, Kristina Nikoli'c et al.

0 citations

2510.21842v22025arXiv

Can Reasoning Models Obfuscate Reasoning? Stress-Testing Chain-of-Thought Monitorability

Artur Zolkowski, Wen Xing, David Lindner et al.

0 citations

2510.19851v12025arXiv

Black-box Optimization of LLM Outputs by Asking for Directions

Jie Zhang, Meng Ding, Yang Liu et al.

0 citations

2510.16794v12025arXiv

The Attacker Moves Second: Stronger Adaptive Attacks Bypass Defenses Against Llm Jailbreaks and Prompt Injections

Milad Nasr, Nicholas Carlini, Chawin Sitawarin et al.

2 citations

2510.09023v12025arXiv

Apertus: Democratizing Open and Compliant LLMs for Global Language Environments

Project Apertus, Alejandro Hern'andez-Cano, Alexander Hagele et al.

17 citations

2509.14233v22025arXiv

Design Patterns for Securing LLM Agents against Prompt Injections

Luca Beurer-Kellner, Beat Buesser, Ana-Maria Cretu et al.

2 citations

2506.08837v32025arXiv

Membership Inference Attacks on Sequence Models

Lorenzo Rossi, Michael Aerni, Jie Zhang et al.

4 citations

2506.05126v12025arXiv

Pitfalls in Evaluating Language Model Forecasters

Daniel Paleka, Shashwat Goel, Jonas Geiping et al.

0 citations

2506.00723v12025arXiv

RealMath: A Continuous Benchmark for Evaluating Language Models on Research-Level Mathematics

Jie Zhang, Cezara Petrui, Kristina Nikoli'c et al.

2 citations

2505.12575v22025arXiv

LLMs unlock new paths to monetizing exploits

Nicholas Carlini, Milad Nasr, Edoardo Debenedetti et al.

0 citations

2505.11449v12025arXiv

The Jailbreak Tax: How Useful are Your Jailbreak Outputs?

Kristina Nikoli'c, Luze Sun, Jie Zhang et al.

0 citations

2504.10694v12025arXiv

Defeating Prompt Injections by Design

Edoardo Debenedetti, Ilia Shumailov, Tianqi Fan et al.

16 citations

2503.18813v22025arXiv

AutoAdvExBench: Benchmarking autonomous exploitation of adversarial example defenses

Nicholas Carlini, Javier Rando, Edoardo Debenedetti et al.

1 citation

2503.01811v12025arXiv

Position: Adversarial ML for LLMs Is Not Making Any Progress

Javier Rando, Jie Zhang, Nicholas Carlini et al.

1 citation

2502.02260v22025arXiv

International AI Safety Report

Yoshua Bengio, Soren Mindermann, Daniel Privitera et al.

16 citations

2501.17805v12025arXiv

Exploring and Mitigating Adversarial Manipulation of Voting-Based Leaderboards

Yangsibo Huang, Milad Nasr, Anastasios Angelopoulos et al.

0 citations

2501.07493v12025arXiv

Consistency Checks for Language Model Forecasters

Daniel Paleka, Abhimanyu Pallavi Sudhir, Alejandro Alvarez et al.

0 citations

2412.18544v22024arXiv

SoK: Watermarking for AI-Generated Content

Xuandong Zhao, Sam Gunn, Miranda Christ et al.

25 citations

2411.18479v32024arXiv

Evaluating the Robustness of the "Ensemble Everything Everywhere" Defense

Jie Zhang, Christian Schlarmann, Kristina Nikoli'c et al.

0 citations

2411.14834v22024arXiv

Measuring Non-Adversarial Reproduction of Training Data in Large Language Models

Michael Aerni, Javier Rando, Edoardo Debenedetti et al.

0 citations

2411.10242v12024arXiv

International Scientific Report on the Safety of Advanced AI (Interim Report)

Yoshua Bengio, Soren Mindermann, Daniel Privitera et al.

9 citations

2412.05282v22024arXiv

Persistent Pre-Training Poisoning of LLMs

Yiming Zhang, Javier Rando, Ivan Evtimov et al.

2 citations

2410.13722v12024arXiv

Gradient-based Jailbreak Images for Multimodal Fusion Models

Javier Rando, Hannah Korevaar, Erik Brinkman et al.

0 citations

2410.03489v22024arXiv

Membership Inference Attacks Cannot Prove that a Model Was Trained On Your Data

Jie Zhang, Debeshee Das, Gautam Kamath et al.

2 citations

2409.19798v22024arXiv

An Adversarial Perspective on Machine Unlearning for AI Safety

Jakub \Lucki, Boyi Wei, Yangsibo Huang et al.

2 citations

2409.18025v62024arXiv

Extracting Training Data from Document-Based VQA Models

Francesco Pinto, Nathalie Rauschmayr, Florian Tramer et al.

1 citation

2407.08707v12024arXiv

Adversarial Search Engine Optimization for Large Language Models

Fredrik Nestaas, Edoardo Debenedetti, Florian Tramer

1 citation

2406.18382v22024arXiv

Blind Baselines Beat Membership Inference Attacks for Foundation Models

Debeshee Das, Jie Zhang, Florian Tramer

5 citations

2406.16201v22024arXiv

AgentDojo: A Dynamic Environment to Evaluate Prompt Injection Attacks and Defenses for LLM Agents

Edoardo Debenedetti, Jie Zhang, Mislav Balunovi'c et al.

129 citations

2406.13352v32024arXiv

Adversarial Perturbations Cannot Reliably Protect Artists From Generative AI

Robert Honig, Javier Rando, Nicholas Carlini et al.

3 citations

2406.12027v22024arXiv

Dataset and Lessons Learned from the 2024 SaTML LLM Capture-the-Flag Competition

Edoardo Debenedetti, Javier Rando, Daniel Paleka et al.

3 citations

2406.07954v12024arXiv

Evaluations of Machine Learning Privacy Defenses are Misleading

Michael Aerni, Jie Zhang, Florian Tramer

18 citations

2404.17399v22024arXiv

Competition Report: Finding Universal Jailbreak Backdoors in Aligned LLMs

Javier Rando, Francesco Croce, Krystof Mitka et al.

1 citation

2404.14461v22024arXiv

Foundational Challenges in Assuring Alignment and Safety of Large Language Models

Usman Anwar, Abulhair Saparov, Javier Rando et al.

19 citations

2404.09932v22024arXiv

Career total: at least 144 works. 106 are in this corpus.Showing the 50 most recent.

Profile built from the corpus for this byline.

144

works

19.8K

citations

51

h-index

Corpus papers per year

Author records are still filling in while the Hub is in beta. If this is your page, you'll be able to claim it soon. Spot a mistake? Tell us.