Related papers: The Mirror Design Pattern: Strict Data Geometry ov…

Every major benchmark for LLM memory systems, LoCoMo foremost, measures whether a model answered correctly, not whether the memory system retrieved correctly. A system returning its entire belief store achieves recall of 1.0 and passes…

Adversarial attacks present a significant security risk to image recognition tasks. Defending against these attacks in a real-life setting can be compared to the way antivirus software works, with a key consideration being how well the…

The recent emergence of multimodal large language models (LLMs) has introduced new opportunities for improving visual hazard recognition on construction sites. Unlike traditional computer vision models that rely on domain-specific training…

Neural backdoor attack is emerging as a severe security threat to deep learning, while the capability of existing defense methods is limited, especially for complex backdoor triggers. In the work, we explore the space formed by the pixel…

The simulation of detector response is a vital aspect of data analysis in particle physics, but current Monte Carlo methods are computationally expensive. Machine learning methods, which learn a mapping from incident particle to detector…

Chain-of-thought prompting has demonstrated great success in facilitating the reasoning abilities of large language models. In this work, we explore how these enhanced reasoning abilities can be exploited to improve the robustness of large…

Antibodies are versatile proteins that can bind to pathogens and provide effective protection for human body. Recently, deep learning-based computational antibody design has attracted popular attention since it automatically mines the…

Multi-modal large language models (MLLMs) have emerged as powerful tools for analyzing Internet-scale image data, offering significant benefits but also raising critical safety and societal concerns. In particular, open-weight MLLMs may be…

The integration of Large Language Models (LLMs) like GPT-4o into robotic systems represents a significant advancement in embodied artificial intelligence. These models can process multi-modal prompts, enabling them to generate more…

Large language model (LLM)-integrated applications have become increasingly prevalent, yet face critical security vulnerabilities from prompt injection (PI) attacks. Defending against PI attacks faces two major issues: malicious…

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…

Multimodal contrastive learning models (e.g., CLIP) can learn high-quality representations from large-scale image-text datasets, while they exhibit significant vulnerabilities to backdoor attacks, raising serious safety concerns. In this…

Large language models (LLMs) increasingly rely on retrieving information from external corpora. This creates a new attack surface: indirect prompt injection (IPI), where hidden instructions are planted in the corpora and hijack model…

The recent Spectre attacks has demonstrated the fundamental insecurity of current computer microarchitecture. The attacks use features like pipelining, out-of-order and speculation to extract arbitrary information about the memory contents…

LLM-integrated applications are vulnerable to prompt injection attacks, where an attacker contaminates the input to inject malicious instructions, causing the LLM to follow the attacker's intent instead of the original user's. Existing…

Information leakage issues in machine learning-based Web applications have attracted increasing attention. While the risk of data privacy leakage has been rigorously analyzed, the theory of model function leakage, known as Model Extraction…

Many deep neural networks are susceptible to minute perturbations of images that have been carefully crafted to cause misclassification. Ideally, a robust classifier would be immune to small variations in input images, and a number of…

Backdoor attacks are an insidious security threat against machine learning models. Adversaries can manipulate the predictions of compromised models by inserting triggers into the training phase. Various backdoor attacks have been devised…

Utilizing task-invariant knowledge acquired from related tasks as prior information, meta-learning offers a principled approach to learning a new task with limited data records. Sample-efficient adaptation of this prior information is a…

Prompt engineering is a technique that involves augmenting a large pre-trained model with task-specific hints, known as prompts, to adapt the model to new tasks. Prompts can be created manually as natural language instructions or generated…