Soheil Feizi
- Papers
- 21
Cite
Notes
Only stored in your browser.
Authored papers
21Failing to Explore: Language Models on Interactive Tasks
arXiv 2026
Adversarial Paraphrasing: A Universal Attack for Humanizing AI-Generated Text
arXiv 2025
Schoenfeld's Anatomy of Mathematical Reasoning by Language Models
arXiv 2025
Almost AI, Almost Human: The Challenge of Detecting AI-Polished Writing
arXiv 2025
DyePack: Provably Flagging Test Set Contamination in LLMs Using Backdoors
arXiv 2025
Gaming Tool Preferences in Agentic LLMs
arXiv 2025
Loki: Low-rank Keys for Efficient Sparse Attention
arXiv 2024
RESTOR: Knowledge Recovery through Machine Unlearning
arXiv 2024
Understanding and Mitigating Compositional Issues in Text-to-Image Generative Models
arXiv 2024
Fast Adversarial Attacks on Language Models In One GPU Minute
arXiv 2024
What do we learn from inverting CLIP models?
arXiv 2024
Certifying LLM Safety against Adversarial Prompting
arXiv 2023
Can AI-Generated Text be Reliably Detected?
arXiv 2023
CUDA: Convolution-based Unlearnable Datasets
CVPR 2023 1
DRSM: De-Randomized Smoothing on Malware Classifier Providing Certified Robustness
arXiv 2023
Exploring Geometry of Blind Spots in Vision Models
exploring-geometry-of-blind-spots-in-vision
Robustness of AI-Image Detectors: Fundamental Limits and Practical Attacks
arXiv 2023
Text-To-Concept (and Back) via Cross-Model Alignment
arXiv 2023
Run-Off Election: Improved Provable Defense against Data Poisoning Attacks
arXiv 2023
Towards Improved Input Masking for Convolutional Neural Networks
ICCV 2023 1
FOCUS: Familiar Objects in Common and Uncommon Settings
focus-familiar-objects-in-common-and-uncommon-1
Affiliations
Frequent co-authors
10from 21 papers