Search Results for author:
Found 12 papers, 11 papers with code
Generating Potent Poisons and Backdoors from Scratch with Guided Diffusion
As a result, we may be able to craft more potent poisons by carefully choosing the base samples.
Spotting LLMs With Binoculars: Zero-Shot Detection of Machine-Generated Text
Detecting text generated by modern large language models is thought to be hard, as both LLMs and humans can exhibit a wide range of complex behaviors.
NEFTune: Noisy Embeddings Improve Instruction Finetuning
We show that language model finetuning can be improved, sometimes dramatically, with a simple augmentation.
Baseline Defenses for Adversarial Attacks Against Aligned Language Models
We find that the weakness of existing discrete optimizers for text, combined with the relatively high costs of optimization, makes standard adaptive attacks more challenging for LLMs.
Bring Your Own Data! Self-Supervised Evaluation for Large Language Models
With the rise of Large Language Models (LLMs) and their ubiquitous deployment in diverse domains, measuring language model behavior on realistic data is imperative.
Revisiting Image Classifier Training for Improved Certified Robust Defense against Adversarial Patches
The success of this strategy relies heavily on the model's invariance to image pixel masking.
On the Reliability of Watermarks for Large Language Models
We also consider a range of new detection schemes that are sensitive to short spans of watermarked text embedded inside a large document, and we compare the robustness of watermarking to other kinds of detectors.
Backdoor Attacks on Vision Transformers
Vision Transformers (ViT) have recently demonstrated exemplary performance on a variety of vision tasks and are being used as an alternative to CNNs.
Backdoor Attacks on Self-Supervised Learning
We show that such methods are vulnerable to backdoor attacks - where an attacker poisons a small part of the unlabeled data by adding a trigger (image patch chosen by the attacker) to the images.
Hidden Trigger Backdoor Attacks
Backdoor attacks are a form of adversarial attacks on deep networks where the attacker provides poisoned data to the victim to train the model with, and then activates the attack by showing a specific small trigger pattern at the test time.
Role of Spatial Context in Adversarial Robustness for Object Detection
However, one can show that an adversary can design adversarial patches which do not overlap with any objects of interest in the scene and exploit contextual reasoning to fool standard detectors.
Universal Litmus Patterns: Revealing Backdoor Attacks in CNNs
In this paper, we introduce a benchmark technique for detecting backdoor attacks (aka Trojan attacks) on deep convolutional neural networks (CNNs).
