Detecting Semantic Backdoors in a Mystery Shopping Scenario
Detecting semantic backdoors in classification models--where some classes can be activated by certain natural, but out-of-distribution inputs--is an important problem that has received relatively little attention. Semantic backdoors are significantly harder to detect than backdoors that are based on trigger patterns due to the lack of such clearly identifiable patterns. We tackle this problem under the assumption that the clean training dataset and the training recipe of the model are both known. These assumptions are motivated by a consumer protection scenario, in which the responsible authority performs mystery shopping to test a machine learning service provider. In this scenario, the authority uses the provider's resources and tools to train a model on a given dataset and tests whether the provider included a backdoor. In our proposed approach, the authority creates a reference model pool by training a small number of clean and poisoned models using trusted infrastructure, and calibrates a model distance threshold to identify clean models. We propose and experimentally analyze a number of approaches to compute model distances and we also test a scenario where the provider performs an adaptive attack to avoid detection. The most reliable method is based on requesting adversarial training from the provider. The model distance is best measured using a set of input samples generated by inverting the models in such a way as to maximize the distance from clean samples. With these settings, our method can often completely separate clean and poisoned models, and it proves to be superior to state-of-the-art backdoor detectors as well.
Code (0)
등록된 구현이 없습니다.
Similar Papers 제목 키워드 기반
DeBackdoor: A Deductive Framework for Detecting Backdoor Attacks on Deep Models with Limited Data
Backdoor attacks are among the most effective, practical, and stealthy attacks in deep learning. In this paper, we consider a practical scenario where a developer obtains a deep model from a third party and uses it as pa…
Backdoor AttackModel Pairing Using Embedding Translation for Backdoor Attack Detection on Open-Set Classification Tasks
Backdoor attacks allow an attacker to embed a specific vulnerability in a machine learning algorithm, activated when an attacker-chosen pattern is presented, causing a specific misprediction. The need to identify backdoo…
Backdoor Attackopen-set classificationDetecting and Eliminating Neural Network Backdoors Through Active Paths with Application to Intrusion Detection
Machine learning backdoors have the property that the machine learning model should work as expected on normal inputs, but when the input contains a specific $\textit{trigger}$, it behaves as the attacker desires. Detect…
Intrusion DetectionLMSanitator: Defending Prompt-Tuning Against Task-Agnostic Backdoors
Prompt-tuning has emerged as an attractive paradigm for deploying large-scale language models due to its strong downstream task performance and efficient multitask serving ability. Despite its wide adoption, we empirical…
MetaBackdoor: Exploiting Positional Encoding as a Backdoor Attack Surface in LLMs
Backdoor attacks pose a serious security threat to large language models (LLMs), which are increasingly deployed as general-purpose assistants in safety- and privacy-critical applications. Existing LLM backdoors rely pri…