相关论文: Building Production-Ready Probes For Gemini

As large language models (LLMs) advance, ensuring AI safety and alignment is paramount. One popular approach is prompt guards, lightweight mechanisms designed to filter malicious queries while being easy to implement and update. In this…

AI leaders and safety reports increasingly warn that advances in model reasoning may enable biological misuse, including by low-expertise users, while major labs describe safeguards as expanding but still evolving rather than settled. This…

Recent advances in large language models (LLMs) have opened new avenues for accelerating scientific research. While models are increasingly capable of assisting with routine tasks, their ability to contribute to novel, expert-level…

The choice of activation function is an active area of research, with different proposals aimed at improving optimization, while maintaining expressivity. Additionally, the activation function can significantly alter the implicit inductive…

This study evaluates the biases in Gemini 2.0 Flash Experimental, a state-of-the-art large language model (LLM) developed by Google, focusing on content moderation and gender disparities. By comparing its performance to ChatGPT-4o, examined…

Gemini is increasingly used to perform tasks on behalf of users, where function-calling and tool-use capabilities enable the model to access user data. Some tools, however, require access to untrusted data introducing risk. Adversaries can…

Conflict prediction in communication is integral to the design of virtual agents that support successful teamwork by providing timely assistance. The aim of our research is to analyze discourse to predict collaboration success.…

As generative AI systems, including large language models (LLMs) and diffusion models, advance rapidly, their growing adoption has led to new and complex security risks often overlooked in traditional AI risk assessment frameworks. This…

Large language models (LLMs) remain vulnerable to sophisticated prompt engineering attacks that exploit contextual framing to bypass safety mechanisms, posing significant risks in cybersecurity applications. We introduce Jailbreak Mimicry,…

Adversarial attacks can reliably steer safety-aligned large language models toward unsafe behavior. Empirically, we find that adversarial prompt-injection attacks can amplify attack success rate from the slow polynomial growth observed…

Large Language Models (LLMs) are increasingly integrated into high-stakes applications, making robust safety guarantees a central practical and commercial concern. Existing safety evaluations predominantly rely on fixed collections of…

Detecting AI-generated images (AIGI) remains challenging because detectors often fail to generalize to unseen generators. Although existing methods are trained on large datasets, their performance still degrades when generation settings…

Large Language Models remain vulnerable to adversarial prompts that elicit toxic content even after safety alignment. We present ToxSearch, a black-box evolutionary framework that tests model safety by evolving prompts in a synchronous…

Meta-Reinforcement learning approaches aim to develop learning procedures that can adapt quickly to a distribution of tasks with the help of a few examples. Developing efficient exploration strategies capable of finding the most useful…

Monitoring is an important aspect of safely deploying Large Language Models (LLMs). This paper examines activation probes for detecting ``high-stakes'' interactions -- where the text indicates that the interaction might lead to significant…

This work investigates the potential of undermining both fairness and detection performance in abusive language detection. In a dynamic and complex digital world, it is crucial to investigate the vulnerabilities of these detection models to…

With the rapid advancement of large language models (LLMs), ensuring their safe use becomes increasingly critical. Fine-tuning is a widely used method for adapting models to downstream tasks, yet it is vulnerable to jailbreak attacks.…

Modern distributed systems are highly dynamic and scalable, requiring monitoring solutions that can adapt to rapid changes. Monitoring systems that rely on external probes can only achieve adaptation through expensive operations such as…

Scenario mining from extensive autonomous driving datasets, such as Argoverse 2, is crucial for the development and validation of self-driving systems. The RefAV framework represents a promising approach by employing Large Language Models…

Large Language Models increasingly power critical infrastructure from healthcare to finance, yet their vulnerability to adversarial manipulation threatens system integrity and user safety. Despite growing deployment, no comprehensive…