Current Retrieval-Augmented Generation systems use uniform processing, causing inefficiency as simple queries consume resources similar to complex multi-hop tasks. We present SymRAG, a framework that introduces adaptive query routing via real-time complexity and load assessment to select symbolic, neural, or hybrid pathways. SymRAG's neuro-symbolic approach adjusts computational pathways based on both query characteristics and system load, enabling efficient resource allocation across diverse query types. By combining linguistic and structural query properties with system load metrics, SymRAG allocates resources proportional to reasoning requirements. Evaluated on 2,000 queries across HotpotQA (multi-hop reasoning) and DROP (discrete reasoning) using Llama-3.2-3B and Mistral-7B models, SymRAG achieves competitive accuracy (97.6--100.0% exact match) with efficient resource utilization (3.6--6.2% CPU utilization, 0.985--3.165s processing). Disabling adaptive routing increases processing time by 169--1151%, showing its significance for complex models. These results suggest adaptive computation strategies are more sustainable and scalable for hybrid AI systems that use dynamic routing and neuro-symbolic frameworks.

Language models often struggle with cross-mode knowledge retrieval -- the ability to access knowledge learned in one format (mode) when queried in another. We demonstrate that models trained on multiple data sources (e.g., Wikipedia and TinyStories) exhibit significantly reduced accuracy when retrieving knowledge in a format different from its original training mode. This paper quantitatively investigates this phenomenon through a controlled study of random token sequence memorization across different modes. We first explore dataset rewriting as a solution, revealing that effective cross-mode retrieval requires prohibitively extensive rewriting efforts that follow a sigmoid-like relationship. As an alternative, we propose CASCADE, a novel pretraining algorithm that uses cascading datasets with varying sequence lengths and computing losses on only the second half of each training sequence to capture knowledge at different scales. Our experiments demonstrate that CASCADE outperforms dataset rewriting approaches, even when compressed into a single model with a unified loss function. This work provides both qualitative evidence of cross-mode retrieval limitations and a practical solution to enhance language models' ability to access knowledge independently of its presentational format.

ABSTRACT With the rapid advancement of generative audio models, distinguishing between human-composed and generated music is becoming increasingly challenging. As a response, models for detecting fake music have been proposed. In this work, we explore the robustness of such systems under audio augmentations. To evaluate model generalization, we constructed a dataset consisting of both real and synthetic music generated using several systems. We then apply a range of audio transformations and analyze how they affect classification accuracy. We test the performance of a recent state-of-the-art musical deepfake detection model in the presence of audio augmentations.

Tensor Network (TN) decompositions have emerged as an indispensable tool in Big Data analytics owing to their ability to provide compact low-rank representations, thus alleviating the ``Curse of Dimensionality'' inherent in handling higher-order data. At the heart of their success lies the concept of TN ranks, which governs the efficiency and expressivity of TN decompositions. However, unlike matrix ranks, TN ranks often lack a universal meaning and an intuitive interpretation, with their properties varying significantly across different TN structures. Consequently, TN ranks are frequently treated as empirically tuned hyperparameters, rather than as key design parameters inferred from domain knowledge. The aim of this Lecture Note is therefore to demystify the foundational yet frequently misunderstood concept of TN ranks through real-life examples and intuitive visualizations. We begin by illustrating how domain knowledge can guide the selection of TN ranks in widely-used models such as the Canonical Polyadic (CP) and Tucker decompositions. For more complex TN structures, we employ a self-explanatory graphical approach that generalizes to tensors of arbitrary order. Such a perspective naturally reveals the relationship between TN ranks and the corresponding ranks of tensor unfoldings (matrices), thereby circumventing cumbersome multi-index tensor algebra while facilitating domain-informed TN design. It is our hope that this Lecture Note will equip readers with a clear and unified understanding of the concept of TN rank, along with the necessary physical insight and intuition to support the selection, explainability, and deployment of tensor methods in both practical applications and educational contexts.

We revise the definition of clickbait, which lacks current consensus, and argue that the creation of a curiosity gap is the key concept that distinguishes clickbait from other related phenomena such as sensationalism and headlines that do not deliver what they promise or diverge from the article. Therefore, we propose a new definition: clickbait is a technique for generating headlines and teasers that deliberately omit part of the information with the goal of raising the readers' curiosity, capturing their attention and enticing them to click. We introduce a new approach to clickbait detection datasets creation, by refining the concept limits and annotations criteria, minimizing the subjectivity in the decision as much as possible. Following it, we created and release TA1C (for Te Ahorré Un Click, Spanish for Saved You A Click), the first open source dataset for clickbait detection in Spanish. It consists of 3,500 tweets coming from 18 well known media sources, manually annotated and reaching a 0.825 Fleiss' κ inter annotator agreement. We implement strong baselines that achieve 0.84 in F1-score.

Health misinformation during the COVID-19 pandemic has significantly challenged public health efforts globally. This study applies the Elaboration Likelihood Model (ELM) to enhance misinformation detection on social media using a hybrid Convolutional Neural Network (CNN) and Long Short-Term Memory (LSTM) model. The model aims to enhance the detection accuracy and reliability of misinformation classification by integrating ELM-based features such as text readability, sentiment polarity, and heuristic cues (e.g., punctuation frequency). The enhanced model achieved an accuracy of 97.37%, precision of 96.88%, recall of 98.50%, F1-score of 97.41%, and ROC-AUC of 99.50%. A combined model incorporating feature engineering further improved performance, achieving a precision of 98.88%, recall of 99.80%, F1-score of 99.41%, and ROC-AUC of 99.80%. These findings highlight the value of ELM features in improving detection performance, offering valuable contextual information. This study demonstrates the practical application of psychological theories in developing advanced machine learning algorithms to address health misinformation effectively.

Model steering represents a powerful technique that dynamically aligns large language models (LLMs) with human preferences during inference. However, conventional model-steering methods rely heavily on externally annotated data, not only limiting their adaptability to varying contexts but also tethering their effectiveness to annotation quality. In this paper, we present SIMS, the first self-improving model-steering framework that operates without relying on external supervision. At its core, SIMS autonomously generates and refines contrastive samples through iterative self-improvement cycles, enabling adaptive, context-specific steering. Additionally, SIMS employs novel strategies, including prompt ranking and contrast sampling, to further enhance steering efficacy. Extensive evaluation across diverse LLMs and benchmarks demonstrates that SIMS substantially outperforms existing methods in steering effectiveness and adaptability, highlighting self-improving model steering as a promising direction for future research on inference-time LLM alignment.

We present a visualization infrastructure that maps data elements to agents, which have behaviors parameterized by those elements. Dynamic visualizations emerge as the agents change position, alter appearance and respond to one other. Agents move to minimize the difference between displayed agent-to-agent distances, and an input matrix of ideal distances. Our current application is visualization of streaming text. Each agent represents a significant word, visualizing it by displaying the word itself, centered in a circle sized by the frequency of word occurrence. We derive the ideal distance matrix from word cooccurrence, mapping higher co-occurrence to lower distance. To depict co-occurrence in its textual context, the ratio of intersection to circle area approximates the ratio of word co-occurrence to frequency. A networked backend process gathers articles from news feeds, blogs, Digg or Twitter, exploiting online search APIs to focus on user-chosen topics. Resulting visuals reveal the primary topics in text streams as clusters, with agent-based layout moving without instability as data streams change dynamically.

In an era of rampant misinformation, generating reliable news explanations is vital, especially for under-represented languages like Hindi. Lacking robust automated tools, Hindi faces challenges in scaling misinformation detection. To bridge this gap, we propose a novel framework integrating Direct Preference Optimization (DPO) with curriculum learning to align machine-generated explanations with human reasoning. Fact-checked explanations from credible sources serve as preferred responses, while LLM outputs highlight system limitations and serve as non-preferred responses. To refine task-specific alignment, we introduce two key parameters -- Actuality and Finesse -- into the DPO loss function, enhancing explanation quality and consistency. Experiments with LLMs (Mistral, Llama, Gemma) and PLMs (mBART, mT5) confirm the framework's effectiveness in generating coherent, contextually relevant explanations. This scalable approach combats misinformation and extends automated explanation generation to low-resource languages.

Video-to-audio (V2A) generation shows great potential in fields such as film production. Despite significant advances, current V2A methods relying on global video information struggle with complex scenes and generating audio tailored to specific objects. To address these limitations, we introduce Hear-Your-Click, an interactive V2A framework enabling users to generate sounds for specific objects by clicking on the frame. To achieve this, we propose Object-aware Contrastive Audio-Visual Fine-tuning (OCAV) with a Mask-guided Visual Encoder (MVE) to obtain object-level visual features aligned with audio. Furthermore, we tailor two data augmentation strategies, Random Video Stitching (RVS) and Mask-guided Loudness Modulation (MLM), to enhance the model's sensitivity to segmented objects. To measure audio-visual correspondence, we designed a new evaluation metric, the CAV score. Extensive experiments demonstrate that our framework offers more precise control and improves generation performance across various metrics. Project Page: https://github.com/SynapGrid/Hear-Your-Click