相关论文: Dynamic Adversarial Fine-Tuning Reorganizes Refusa…

Adversarial examples, crafted by adding perturbations imperceptible to humans, can deceive neural networks. Recent studies identify the adversarial transferability across various models, \textit{i.e.}, the cross-model attack ability of…

Deep reinforcement learning models are vulnerable to adversarial attacks that can decrease a victim's cumulative expected reward by manipulating the victim's observations. Despite the efficiency of previous optimization-based methods for…

Unified multi-modal encoders that bind vision, audio, and other sensors into a shared embedding space are attractive building blocks for robot perception and decision-making. However, on-robot deployment exposes the vision branch to…

Instruction Fine-Tuning (IFT) has been widely adopted as an effective post-training strategy to enhance various abilities of Large Language Models (LLMs). However, prior studies have shown that IFT can significantly compromise LLMs' safety,…

Recently, the field of adversarial machine learning has been garnering attention by showing that state-of-the-art deep neural networks are vulnerable to adversarial examples, stemming from small perturbations being added to the input image.…

Recently, adversarial training has been incorporated in self-supervised contrastive pre-training to augment label efficiency with exciting adversarial robustness. However, the robustness came at a cost of expensive adversarial training. In…

With the evolution of self-supervised learning, the pre-training paradigm has emerged as a predominant solution within the deep learning landscape. Model providers furnish pre-trained encoders designed to function as versatile feature…

Safety-aligned language models are trained to refuse harmful requests, yet refusal behavior can be suppressed by steering their internal representations. Existing methods do so by ablating a refusal direction from model activations, aiming…

The performance of state-of-the-art neural rankers can deteriorate substantially when exposed to noisy inputs or applied to a new domain. In this paper, we present a novel method for fine-tuning neural rankers that can significantly improve…

Large kernel convolutions offer a scalable alternative to vision transformers for high-resolution 3D volumetric analysis, yet naively increasing kernel size often leads to optimization instability. Motivated by the spatial bias inherent in…

Activation steering has emerged as a promising approach for efficiently adapting large language models (LLMs) to downstream behaviors. However, most existing steering methods rely on a single static direction per task or concept, making…

Single-turn safety evaluation is a poor proxy for real fraud defense, where attackers escalate across multiple rounds. This paper evaluates fraud defenders under replay and adaptive multi-round attacks and measures when a defender refuses,…

With the rise of powerful foundation models, a pre-training-fine-tuning paradigm becomes increasingly popular these days: A foundation model is pre-trained using a huge amount of data from various sources, and then the downstream users only…

Deep neural models (e.g. Transformer) naturally learn spurious features, which create a ``shortcut'' between the labels and inputs, thus impairing the generalization and robustness. This paper advances the self-attention mechanism to its…

Adversarial robustness has been studied extensively in image classification, especially for the $\ell_\infty$-threat model, but significantly less so for related tasks such as object detection and semantic segmentation, where attacks turn…

Transformer based large language models with emergent capabilities are becoming increasingly ubiquitous in society. However, the task of understanding and interpreting their internal workings, in the context of adversarial attacks, remains…

Recent works have shown theoretically and empirically that redundant data dimensions are a source of adversarial vulnerability. However, the inverse doesn't seem to hold in practice; employing dimension-reduction techniques doesn't exhibit…

Visual language models (VLMs) have made significant progress in image captioning tasks, yet recent studies have found they are vulnerable to backdoor attacks. Attackers can inject undetectable perturbations into the data during inference,…

Leading language model (LM) providers like OpenAI and Anthropic allow customers to fine-tune frontier LMs for specific use cases. To prevent abuse, these providers apply filters to block fine-tuning on overtly harmful data. In this setting,…

Reinforcement fine-tuning (RFT) has become a core paradigm for post-training large language models, yet its training process remains highly fragile. Existing efforts mainly improve reliability at the system level or address specific issues…