Search Results for author:
Found 11 papers, 6 papers with code
SoK: Unintended Interactions among Machine Learning Defenses and Risks
We survey existing literature on unintended interactions, accommodating them within our framework.
LLM Self Defense: By Self Examination, LLMs Know They Are Being Tricked
We test LLM Self Defense on GPT 3. 5 and Llama 2, two of the current most prominent LLMs against various types of attacks, such as forcefully inducing affirmative responses to prompts and prompt engineering attacks.
False Claims against Model Ownership Resolution
Our core idea is that a malicious accuser can deviate (without detection) from the specified MOR process by finding (transferable) adversarial examples that successfully serve as evidence against independent suspect models.
On the Robustness of Dataset Inference
However, in a subspace of the same setting, we prove that DI suffers from high false positives (FPs) -- it can incorrectly identify an independent model trained with non-overlapping data from the same distribution as stolen.
Conflicting Interactions Among Protection Mechanisms for Machine Learning Models
We then focus on systematically analyzing pairwise interactions between protection mechanisms for one concern, model and data ownership verification, with two other classes of ML protection mechanisms: differentially private training, and robustness against model evasion.
SHAPr: An Efficient and Versatile Membership Privacy Risk Metric for Machine Learning
Using ten benchmark datasets, we show that SHAPr is indeed effective in estimating susceptibility of training data records to MIAs.
Good Artists Copy, Great Artists Steal: Model Extraction Attacks Against Image Translation Models
We present a framework for conducting such attacks, and show that an adversary can successfully extract functional surrogate models by querying $F_V$ using data from the same domain as the training data for $F_V$.
Extraction of Complex DNN Models: Real Threat or Boogeyman?
However, model extraction attacks can steal the functionality of ML models using the information leaked to clients through the results returned via the API.
Detecting organized eCommerce fraud using scalable categorical clustering
Our approach is based on clustering and aims to group together fraudulent orders placed by the same group of fraudsters.
DAWN: Dynamic Adversarial Watermarking of Neural Networks
Existing watermarking schemes are ineffective against IP theft via model extraction since it is the adversary who trains the surrogate model.
PRADA: Protecting against DNN Model Stealing Attacks
Access to the model can be restricted to be only via well-defined prediction APIs.
Cryptography and Security
