Large Language Models (LLMs) are vulnerable to jailbreak attacks, which use crafted prompts to elicit toxic responses. These attacks exploit LLMs' difficulty in dynamically detecting harmful intents during the generation process. Traditional safety alignment methods, often relying on the initial few generation steps, are ineffective due to limited computational budget. This paper proposes DEEPALIGN, a robust defense framework that fine-tunes LLMs to progressively detoxify generated content, significantly improving both the computational budget and effectiveness of mitigating harmful generation. Our approach uses a hybrid loss function operating on hidden states to directly improve LLMs' inherent awareness of toxity during generation. Furthermore, we redefine safe responses by generating semantically relevant answers to harmful queries, thereby increasing robustness against representation-mutation attacks. Evaluations across multiple LLMs demonstrate state-of-the-art defense performance against six different attack types, reducing Attack Success Rates by up to two orders of magnitude compared to previous state-of-the-art defense while preserving utility. This work advances LLM safety by addressing limitations of conventional alignment through dynamic, context-aware mitigation.

In this paper, we address the enhancement of classification accuracy for 3D point cloud Lidar data, an optical remote sensing technique that estimates the three-dimensional coordinates of a given terrain. Our approach introduces product coefficients, theoretical quantities derived from measure theory, as additional features in the classification process. We define and present the formulation of these product coefficients and conduct a comparative study, using them alongside principal component analysis (PCA) as feature inputs. Results demonstrate that incorporating product coefficients into the feature set significantly improves classification accuracy within this new framework.

Transcription of aviation communications has several applications, from assisting air traffic controllers in identifying the accuracy of read-back errors to search and rescue operations. Recent advances in artificial intelligence have provided unprecedented opportunities for improving aviation communication transcription tasks. OpenAI's Whisper is one of the leading automatic speech recognition models. However, fine-tuning Whisper for aviation communication transcription is not computationally efficient. Thus, this paper aims to use a Parameter-Efficient Fine-tuning method called Low-Rank Adaptation to fine-tune a more computationally efficient version of Whisper, distil-Whisper. To perform the fine-tuning, we used the Air Traffic Control Corpus dataset from the Linguistic Data Consortium, which contains approximately 70 hours of controller and pilot transmissions near three major airports in the US. The objective was to reduce the word error rate to enhance accuracy in the transcription of aviation communication. First, starting with an initial set of hyperparameters for LoRA (Alpha = 64 and Rank = 32), we performed a grid search. We applied a 5-fold cross-validation to find the best combination of distil-Whisper hyperparameters. Then, we fine-tuned the model for LoRA hyperparameters, achieving an impressive average word error rate of 3.86% across five folds. This result highlights the model's potential for use in the cockpit.

Phishing attacks continue to impose a significant threat on digital communication and online transactions, costing organizations and individuals billions of dollars each year. According to the Anti-Phishing Working Group (APWG), phishing incidents increased by over 25% in 2022 compared to previous years, with attackers refining their methods to mimic trusted brands and deceive users into revealing sensitive information Anti-Phishing Working Group [2022]. This alarming increase not only highlights the ingenuity of cybercriminals but also emphasizes the critical need for more advanced detection systems. In response, researchers and cybersecurity professionals have increasingly turned to artificial intelligence (AI) and deep learning (DL) techniques to build more accurate and adaptable detection systems capable of identifying subtle cues in phishing attempts. Historically, phishing detection relied on signature-based methods and blacklists, which, although useful, could not keep pace with the rapid evolution of phishing tactics. Traditional approaches often suffered from high false-positive rates and were unable to adapt to new, previously unseen attack vectors.

The choice of neural network features can have a large impact on both the accuracy and speed of the network. Despite the current industry shift towards large transformer models, specialized binary classifiers remain critical for numerous practical applications where computational efficiency and low latency are essential. Neural network features tend to be developed homogeneously, resulting in slower or less accurate networks when testing against multiple datasets. In this paper, we show the effectiveness of Neuvo NAS+ a novel Python implementation of an extended Neural Architecture Search (NAS+) which allows the user to optimise the training parameters of a network as well as the network's architecture. We provide an in-depth analysis of the importance of catering a network's architecture to each dataset. We also describe the design of the Neuvo NAS+ system that selects network features on a task-specific basis including network training hyper-parameters such as the number of epochs and batch size. Results show that the Neuvo NAS+ task-specific approach significantly outperforms several machine learning approaches such as Naive Bayes, C4.5, Support Vector Machine and a standard Artificial Neural Network for solving a range of binary classification problems in terms of accuracy. Our experiments demonstrate substantial diversity in evolved network architectures across different datasets, confirming the value of task-specific optimization. Additionally, Neuvo NAS+ outperforms other evolutionary algorithm optimisers in terms of both accuracy and computational efficiency, showing that properly optimized binary classifiers can match or exceed the performance of more complex models while requiring significantly fewer computational resources.

High-quality test items are essential for educational assessments, particularly within Item Response Theory (IRT). Traditional validation methods rely on resource-intensive pilot testing to estimate item difficulty and discrimination. More recently, Item-Writing Flaw (IWF) rubrics emerged as a domain-general approach for evaluating test items based on textual features. However, their relationship to IRT parameters remains underexplored. To address this gap, we conducted a study involving over 7,000 multiple-choice questions across various STEM subjects (e.g., math and biology). Using an automated approach, we annotated each question with a 19-criteria IWF rubric and studied relationships to data-driven IRT parameters. Our analysis revealed statistically significant links between the number of IWFs and IRT difficulty and discrimination parameters, particularly in life and physical science domains. We further observed how specific IWF criteria can impact item quality more and less severely (e.g., negative wording vs. implausible distractors). Overall, while IWFs are useful for predicting IRT parameters--particularly for screening low-difficulty MCQs--they cannot replace traditional data-driven validation methods. Our findings highlight the need for further research on domain-general evaluation rubrics and algorithms that understand domain-specific content for robust item validation.

In this study, we present an approach for efficient spatiotemporal feature extraction using MobileNetV4 and a multi-scale 3D MLP-Mixer-based temporal aggregation module. MobileNetV4, with its Universal Inverted Bottleneck (UIB) blocks, serves as the backbone for extracting hierarchical feature representations from input image sequences, ensuring both computational efficiency and rich semantic encoding. To capture temporal dependencies, we introduce a three-level MLP-Mixer module, which processes spatial features at multiple resolutions while maintaining structural integrity. Experimental results on the ABAW 8th competition demonstrate the effectiveness of our approach, showing promising performance in affective behavior analysis. By integrating an efficient vision backbone with a structured temporal modeling mechanism, the proposed framework achieves a balance between computational efficiency and predictive accuracy, making it well-suited for real-time applications in mobile and embedded computing environments.

Out-of-distribution (OOD) detection remains challenging for deep learning models, particularly when test-time OOD samples differ significantly from training outliers. We propose OODD, a novel test-time OOD detection method that dynamically maintains and updates an OOD dictionary without fine-tuning. Our approach leverages a priority queue-based dictionary that accumulates representative OOD features during testing, combined with an informative inlier sampling strategy for in-distribution (ID) samples. To ensure stable performance during early testing, we propose a dual OOD stabilization mechanism that leverages strategically generated outliers derived from ID data. To our best knowledge, extensive experiments on the OpenOOD benchmark demonstrate that OODD significantly outperforms existing methods, achieving a 26.0% improvement in FPR95 on CIFAR-100 Far OOD detection compared to the state-of-the-art approach. Furthermore, we present an optimized variant of the KNN-based OOD detection framework that achieves a 3x speedup while maintaining detection performance.

Social media platforms are becoming the foundations of social interactions including messaging and opinion expression. In this regard, Sentiment Analysis techniques focus on providing solutions to ensure the retrieval and analysis of generated data including sentiments, emotions, and discussed topics. International competitions such as the International Workshop on Semantic Evaluation (SemEval) have attracted many researchers and practitioners with a special research interest in building sentiment analysis systems. In our work, we study top-ranking systems for each SemEval edition during the 2013-2021 period, a total of 658 teams participated in these editions with increasing interest over years. We analyze the proposed systems marking the evolution of research trends with a focus on the main components of sentiment analysis systems including data acquisition, preprocessing, and classification. Our study shows an active use of preprocessing techniques, an evolution of features engineering and word representation from lexicon-based approaches to word embeddings, and the dominance of neural networks and transformers over the classification phase fostering the use of ready-to-use models. Moreover, we provide researchers with insights based on experimented systems which will allow rapid prototyping of new systems and help practitioners build for future SemEval editions.

Speaker identification in multilingual settings presents unique challenges, particularly when conventional models are predominantly trained on English data. In this paper, we propose WSI (Whisper Speaker Identification), a framework that repurposes the encoder of the Whisper automatic speech recognition model pre trained on extensive multilingual data to generate robust speaker embeddings via a joint loss optimization strategy that leverages online hard triplet mining and self supervised Normalized Temperature-scaled Cross Entropy loss. By capitalizing on Whisper language-agnostic acoustic representations, our approach effectively distinguishes speakers across diverse languages and recording conditions. Extensive evaluations on multiple corpora, including VoxTube (multilingual), JVS (Japanese), CallHome (German, Spanish, Chinese, and Japanese), and Voxconverse (English), demonstrate that WSI consistently outperforms state-of-the-art baselines, namely Pyannote Embedding, ECAPA TDNN, and Xvector, in terms of lower equal error rates and higher AUC scores. These results validate our hypothesis that a multilingual pre-trained ASR encoder, combined with joint loss optimization, substantially improves speaker identification performance in non-English languages.