Related papers: Differentiable Faithfulness Alignment for Cross-Mo…

Attacks on Federated Learning (FL) can severely reduce the quality of the generated models and limit the usefulness of this emerging learning paradigm that enables on-premise decentralized learning. However, existing untargeted attacks are…

Large Language Models (LLMs) have demonstrated remarkable efficiency in tackling various tasks based on human instructions, but studies reveal that they often struggle with tasks requiring reasoning, such as math or physics. This limitation…

Existing circuit discovery methods rely on templated tasks with clean counterfactuals, limiting their use on diverse natural text. We adapt Contextual Decomposition for Transformers (CD-T) for unstructured settings via label-balanced…

Domain adaptation deals with training models using large scale labeled data from a specific source domain and then adapting the knowledge to certain target domains that have few or no labels. Many prior works learn domain agnostic feature…

Fine-tuning significantly improves the performance of Large Language Models (LLMs), yet its underlying mechanisms remain poorly understood. This paper aims to provide an in-depth interpretation of the fine-tuning process through circuit…

The feedback alignment (FA) algorithm offers a biologically plausible alternative to backpropagation (BP) for training neural networks yet notably fails to scale to convolutional architectures. Modifications have been proposed to address…

This paper presents an unsupervised deep-learning framework named Local Deep-Feature Alignment (LDFA) for dimension reduction. We construct neighbourhood for each data sample and learn a local Stacked Contractive Auto-encoder (SCAE) from…

Adversarial training has been widely acknowledged as the most effective method to improve the adversarial robustness against adversarial examples for Deep Neural Networks (DNNs). So far, most existing works focus on enhancing the overall…

We localize the policy routing mechanism in alignment-trained language models. An intermediate-layer attention gate reads detected content and triggers deeper amplifier heads that boost the signal toward refusal. In smaller models the gate…

We discuss the problem of learning a deterministic finite automaton (DFA) from a confidence oracle. That is, we are given access to an oracle $Q$ with incomplete knowledge of some target language $L$ over an alphabet $\Sigma$; the oracle…

Transformer-based language models excel at both recall (retrieving memorized facts) and reasoning (performing multi-step inference), but whether these abilities rely on distinct internal mechanisms remains unclear. Distinguishing recall…

Existing deepfake analysis methods are primarily based on discriminative models, which significantly limit their application scenarios. This paper aims to explore interactive deepfake analysis by performing instruction tuning on multi-modal…

We examine the Detrended Fluctuation Analysis (DFA), which is a well-established method for the detection of long-range correlations in time series. We show that deviations from scaling that appear at small time scales become stronger in…

Deep learning models continue to advance in accuracy, yet they remain vulnerable to adversarial attacks, which often lead to the misclassification of adversarial examples. Adversarial training is used to mitigate this problem by increasing…

Transformer networks have achieved remarkable success across diverse domains, leveraging a variety of architectural innovations, including residual connections. However, traditional residual connections, which simply sum the outputs of…

For a long time, anomaly localization has been widely used in industries. Previous studies focused on approximating the distribution of normal features without adaptation to a target dataset. However, since anomaly localization should…

In this paper, we study the task of source-free domain adaptation (SFDA), where the source data are not available during target adaptation. Previous works on SFDA mainly focus on aligning the cross-domain distributions. However, they ignore…

While error backpropagation (BP) has dominated the training of nearly all modern neural networks for a long time, it suffers from several biological plausibility issues such as the symmetric weight requirement and synchronous updates.…

With the rapid advancement of large language models (LLMs), their deployment in real-world applications has become increasingly widespread. LLMs are expected to deliver robust performance across diverse tasks, user preferences, and…

In recent years, an increasing popularity of deep learning model for intelligent condition monitoring and diagnosis as well as prognostics used for mechanical systems and structures has been observed. In the previous studies, however, a…