SafetyNet: Detecting Harmful Outputs in LLMs by Modeling and Monitoring Deceptive Behaviors
High-risk industries like nuclear and aviation use real-time monitoring to detect dangerous system conditions. Similarly, Large Language Models (LLMs) need monitoring safeguards. We propose a real-time framework to predict harmful AI outputs before they occur by using an unsupervised approach that treats normal behavior as the baseline and harmful outputs as outliers. Our study focuses specifically on backdoor-triggered responses -- where specific input phrases activate hidden vulnerabilities causing the model to generate unsafe content like violence, pornography, or hate speech. We address two key challenges: (1) identifying true causal indicators rather than surface correlations, and (2) preventing advanced models from deception -- deliberately evading monitoring systems. Hence, we approach this problem from an unsupervised lens by drawing parallels to human deception: just as humans exhibit physical indicators while lying, we investigate whether LLMs display distinct internal behavioral signatures when generating harmful content. Our study addresses two critical challenges: 1) designing monitoring systems that capture true causal indicators rather than superficial correlations; and 2)preventing intentional evasion by increasingly capable "Future models''. Our findings show that models can produce harmful content through causal mechanisms and can become deceptive by: (a) alternating between linear and non-linear representations, and (b) modifying feature relationships. To counter this, we developed Safety-Net -- a multi-detector framework that monitors different representation dimensions, successfully detecting harmful behavior even when information is shifted across representational spaces to evade individual monitors. Our evaluation shows 96% accuracy in detecting harmful cases using our unsupervised ensemble approach.
Code (0)
등록된 구현이 없습니다.
Similar Papers 제목 키워드 기반
SafetyNet: Detecting and Rejecting Adversarial Examples Robustly
We describe a method to produce a network where current methods such as DeepFool have great difficulty producing adversarial samples. Our construction suggests some insights into how deep networks work. We provide a reas…
SafetyNets: Verifiable Execution of Deep Neural Networks on an Untrusted Cloud
Inference using deep neural networks is often outsourced to the cloud since it is a computationally demanding task. However, this raises a fundamental issue of trust. How can a client be sure that the cloud has performed…
speech-recognitionSpeech RecognitionCan We Locate and Prevent Stereotypes in LLMs?
Stereotypes in large language models (LLMs) can perpetuate harmful societal biases. Despite the widespread use of models, little is known about where these biases reside in the neural network. This study investigates the…
RigorLLM: Resilient Guardrails for Large Language Models against Undesired Content
Recent advancements in Large Language Models (LLMs) have showcased remarkable capabilities across various tasks in different domains. However, the emergence of biases and the potential for generating harmful content in L…
Data AugmentationDetecting and Filtering Unsafe Training Data via Data Attribution
Large language models (LLMs) are vulnerable to unsafe training data that even small amounts of unsafe data can lead to harmful model behaviors. Detecting and filtering such unsafe training data is essential for trustwort…