Abstract
Over the last three years, the JOKER Lab series at CLEF has gathered an active community of researchers in natural language processing and information retrieval to collaborate on non-literal use of language in text. Such language can be a challenge for AI systems, but also sometimes for humans, as it requires understanding implicit cultural references and unorthodox interactions between form and meaning. In this paper, we discuss the lessons learned from the previous iterations of the Lab and describe how its upcoming edition will build upon those to address new challenges. In 2025, JOKER will provide novel tasks and update some previous ones with new data and new languages. This year we provide sandbox environments for experimenting with humour-aware information retrieval (Task 1), a previously featured task now enhanced with an all-new Portuguese corpus; wordplay translation in text (Task 2), another historical task for which we provide new corpora; onomastic wordplay (Task 3), a new task focussed on humorous proper names in fiction; and controlled creativity (Task 4), another novel task that aims at identifying and avoiding hallucinations.

Abstract
The increasing penetration of renewable energy sources in power systems has heightened the importance of grid-forming (GFM) converters, which emulate the dynamic behavior of synchronous machines and are crucial for ensuring stability in converter-dominated grids. However, the complexity of modern grids calls for innovative control mechanisms to unlock the full potential of GFM technology. This work presents a novel automated framework for control design in power systems. Simulated annealing is used to evolve the structural design of control systems represented as graph-based models. The method achieves greater flexibility by using control graphs instead of traditional tree-based representations, supporting complex feedback loop configurations. A simplification process is also included to reduce complexity and improve interpretability, ensuring practical applicability. Validation on a two-generator power system with one GFM converter demonstrates the method's ability to design robust controllers that enhance system stability, achieving better performance metrics, such as smoother frequency responses with significantly reduced frequency deviations compared to benchmark configurations. The improved frequency response arises from differing terminal angle profiles, enabling faster, stronger power responses that quickly arrest frequency deviations during disturbances.

Abstract
There was an error in the original publication [1]. The statement in the Acknowledgments section is incorrect and should be removed because the official start of the project WATSON was after the paper’s publication date. The authors state that the scientific conclusions are unaffected. This correction was approved by the Academic Editor. The original publication has also been updated. © 2025 by the authors.

Abstract
Advancements in deep learning have revolutionized materials discovery by enabling predictive modeling of complex material properties. However, single-modal approaches often fail to capture the intricate interplay of compositional, structural, and morphological characteristics. This study introduces a novel multimodal deep learning framework for enhanced material property prediction, integrating textual (chemical compositions), tabular (structural descriptors), and image-based (2D crystal structure visualizations) modalities. Utilizing the Alexandriadatabase, we construct a comprehensive multimodal dataset of 10,000 materials with symmetry-resolved crystallographic data. Specialized neural architectures, such as FT-Transformer for tabular data, Hugging Face Electra-based model for text, and TIMM-based MetaFormer for images, generate modality-specific embeddings, fused through a hybrid strategy into a unified latent space. The framework predicts seven critical material properties, including electronic (band gap, density of states), thermodynamic (formation energy, energy above hull, total energy), magnetic (magnetic moment per volume), and volumetric (volume per atom) features, many governed by crystallographic symmetry. Experimental results demonstrated that multimodal fusion significantly outperforms unimodal baselines. Notably, the bimodal integration of image and text data showed significant gains, reducing the Mean Absolute Error for band gap by approximately 22.7% and for volume per atom by 22.4% compared to the average unimodal models. This combination also achieved a 28.4% reduction in Root Mean Squared Error for formation energy. The full trimodal model (tabular + images + text) yielded competitive, and in several cases the lowest, error metrics, particularly for band gap, magnetic moment per volume and density of states per atom, confirming the value of integrating all three modalities. This scalable, modular framework advances materials informatics, offering a powerful tool for data-driven materials discovery and design.

Abstract
In the age of rapidly advancing machine learning capabilities, the pursuit of maximum performance encounters the practical limitations imposed by limited resources in several fields. This work presents a cost-effective proposal for feature selection, which is a crucial part of machine learning processes, and intends to partly solve this problem through computational time reduction. The proposed methodology aims to strike a careful balance between feature exploration and strict computational time concerns, by enhancing the quality and relevance of data. This approach focuses on the use of interim representations of feature combinations to significantly speed up a potentially slow and computationally expensive process. This strategy is evaluated in several datasets against other feature selection methods, and the results indicate a significant reduction in the temporal costs associated with this process, achieving a mean percentage decrease of 85%. Furthermore, this reduction is achieved while maintaining competitive model performance, demonstrating that the selected features remain effective for the learning task. These results emphasize the method's feasibility, confirming its ability to transform machine learning applications in environments with limited resources.

Abstract
This position paper presents A4FN, an Agentic Artificial Intelligence (AI) architecture for intent-driven automation in Flying Networks (FNs) using Unmanned Aerial Vehicles (UAVs) as access nodes. A4FN leverages Generative AI and Large Language Models (LLMs) to enable real-time, context-aware network control via a distributed agentic system. It comprises two components: the Perception Agent (PA), which semantically interprets multimodal input - including imagery, audio, and telemetry data - from UAV-mounted sensors to derive Service Level Specifications (SLSs); and the Decision-and-Action Agent (DAA), which reconfigures the network based on inferred intents. A4FN embodies key properties of Agentic AI, including autonomy, goal-driven reasoning, and continuous perception-action cycles. Designed for mission-critical, infrastructure-limited scenarios such as disaster response, it supports adaptive reconfiguration, dynamic resource management, and interoperability with emerging wireless technologies. The paper details the A4FN architecture, its core innovations, and open research challenges in multi-agent coordination and Agentic AI integration in next-generation FNs. © 2025 IEEE.