Get started
Author

David Dobre

Soft Prompt Threats: Attacking Safety Alignment and Unlearning in Open-Source LLMs through the Embedding Space
2024 · Leo Schwinn, David Dobre, et al. · neural information processing systems · 57
PDF
Learning diverse attacks on large language models for robust red-teaming and safety tuning
2024 · Lee, Seanie, Kim, Minsu, et al. · arXiv (Cornell University) · 48
PDF
Adversarial Attacks and Defenses in Large Language Models: Old and New Threats
2023 · Schwinn, Leo, Dobre, David, et al. · arXiv (Cornell University) · 44
PDF
In-Context Learning Can Re-learn Forbidden Tasks
2024 · Xhonneux, Sophie, Dobre, David, et al. · arXiv (Cornell University) · 12
PDF
Clipped Stochastic Methods for Variational Inequalities with Heavy-Tailed Noise
2022 · Eduard Gorbunov, Marina Danilova, et al. · neural information processing systems · 12
PDF
Sarah Frank-Wolfe: Methods for Constrained Optimization with Best Rates and Practical Features
2023 · Beznosikov, Aleksandr, Dobre, David, et al. · arXiv (Cornell University) · 9
PDF
Raising the Bar for Certified Adversarial Robustness with Diffusion Models
2023 · Altstidl, Thomas, Dobre, David, et al. · arXiv (Cornell University) · 8
PDF
Dissecting adaptive methods in GANs
2022 · Jelassi, Samy, Dobre, David, et al. · arXiv (Cornell University) · 6
PDF
On the Scalability of Certified Adversarial Robustness with Generated Data
2024 · Thomas Altstidl, David Dobre, et al. · neural information processing systems · 4
A Generative Approach to LLM Harmfulness Mitigation with Red Flag Tokens
2025 · Dobre, David, Mofakhami, Mehrnaz, et al. · arXiv (Cornell University) · 3
PDF
Visual Memory Attacks Can Persist Through The KV Cache
2026 · Dobre, David, Schwinn, Leo, et al. · arXiv (Cornell University) · 1
PDF
Works count: 11
Citation count: 81
H-index: 4
I10-index: 3
Year
Topic
Type