Discovering functional crystalline materials entails navigating an immense combinatorial design space. While recent advances in generative artificial intelligence have enabled the sampling of chemically plausible compositions and structures, a fundamental challenge remains: the objective misalignment between likelihood-based sampling in generative modelling and targeted focus on underexplored regions where novel compounds reside. Here, we introduce a reinforcement learning framework that guides latent denoising diffusion models toward diverse and novel, yet thermodynamically viable crystalline compounds. Our approach integrates group relative policy optimisation with verifiable, multi-objective rewards that jointly balance creativity, stability, and diversity. Beyond de novo generation, we demonstrate enhanced property-guided design that preserves chemical validity, while targeting desired functional properties. This approach establishes a modular foundation for controllable AI-driven inverse design that addresses the novelty-validity trade-off across scientific discovery applications of generative models.

The rapid expansion of Large Language Models (LLMs) has introduced unprecedented energy demands, extending beyond training to large-scale inference workloads that often dominate total lifecycle consumption. Deploying these models requires energy-intensive GPU infrastructure, and in some cases has even prompted plans to power data centers with nuclear energy. Despite this growing relevance, systematic analyses of inference energy consumption remain limited. In this work, we present a large-scale measurement-based study comprising over 32,500 measurements across 21 GPU configurations and 155 model architectures, from small open-source models to frontier systems. Using the vLLM inference engine, we quantify energy usage at the prompt level and identify how architectural and operational factors shape energy demand. Building on these insights, we develop a predictive model that accurately estimates inference energy consumption across unseen architectures and hardware, and implement it as a browser extension to raise awareness of the environmental impact of generative AI.

Benchmarks play a significant role in how researchers and the public understand generative AI systems. However, the widespread use of benchmark scores to communicate about model capabilities has led to criticisms of validity, especially whether benchmarks test what they claim to test (i.e. construct validity) and whether benchmark evaluations are representative of how models are used in the wild (i.e. ecological validity). In this work we explore how to create an LLM benchmark that addresses these issues by taking a human-centered approach. We focus on designing a domain-oriented benchmark for journalism practitioners, drawing on insights from a workshop of 23 journalism professionals. Our workshop findings surface specific challenges that inform benchmark design opportunities, which we instantiate in a case study that addresses underlying criticisms and specific domain concerns. Through our findings and design case study, this work provides design guidance for developing benchmarks that are better tuned to specific domains.

Secure data hiding remains a fundamental challenge in digital communication, requiring a careful balance between computational efficiency and perceptual transparency. The balance between security and performance is increasingly fragile with the emergence of generative AI systems capable of autonomously generating and optimising sophisticated cryptanalysis and steganalysis algorithms, thereby accelerating the exposure of vulnerabilities in conventional data-hiding schemes. This work introduces SteganoSNN, a neuromorphic steganographic framework that exploits spiking neural networks (SNNs) to achieve secure, low-power, and high-capacity multimedia data hiding. Digitised audio samples are converted into spike trains using leaky integrate-and-fire (LIF) neurons, encrypted via a modulo-based mapping scheme, and embedded into the least significant bits of RGBA image channels using a dithering mechanism to minimise perceptual distortion. Implemented in Python using NEST and realised on a PYNQ-Z2 FPGA, SteganoSNN attains real-time operation with an embedding capacity of 8 bits per pixel. Experimental evaluations on the DIV2K 2017 dataset demonstrate image fidelity between 40.4 dB and 41.35 dB in PSNR and SSIM values consistently above 0.97, surpassing SteganoGAN in computational efficiency and robustness. SteganoSNN establishes a foundation for neuromorphic steganography, enabling secure, energy-efficient communication for Edge-AI, IoT, and biomedical applications.

This article presents two corpora of English and Czech texts generated with large language models (LLMs). The motivation is to create a resource for comparing human-written texts with LLM-generated text linguistically. Emphasis was placed on ensuring these resources are multi-genre and rich in terms of topics, authors, and text types, while maintaining comparability with existing human-created corpora. These generated corpora replicate reference human corpora: BE21 by Paul Baker, which is a modern version of the original Brown Corpus, and Koditex corpus that also follows the Brown Corpus tradition but in Czech. The new corpora were generated using models from OpenAI, Anthropic, Alphabet, Meta, and DeepSeek, ranging from GPT-3 (davinci-002) to GPT-4.5, and are tagged according to the Universal Dependencies standard (i.e., they are tokenized, lemmatized, and morphologically and syntactically annotated). The subcorpus size varies according to the model used (the English part contains on average 864k tokens per model, 27M tokens altogether, the Czech partcontains on average 768k tokens per model, 21.5M tokens altogether). The corpora are freely available for download under the CC BY 4.0 license (the annotated data are under CC BY-NC-SA 4.0 licence) and are also accessible through the search interface of the Czech National Corpus.

--Swarm intelligence describes how simple, decentralized agents can collectively produce complex behaviors. Recently, the concept of swarming has been extended to large language model (LLM)-powered systems, such as OpenAI's Swarm (OAS) framework, where agents coordinate through natural language prompts. Using OAS, we implement and compare classical and LLMbased versions of two well-established swarm algorithms: Boids and Ant Colony Optimization. Results indicate that while LLMpowered swarms can emulate swarm-like dynamics, they are constrained by substantial computational overhead. For instance, our LLM-based Boids simulation required roughly 300 more computation time than its classical counterpart, highlighting current limitations in applying LLM-driven swarms to real-time systems. W ARM intelligence continues to attract significant attention from researchers and engineers. In nature, swarming systems exist as flocks of birds, schools of fish, and colonies of ants, where they are characterized by local interactions among agents following simple rules. These interactions give rise to global patterns and adaptive behaviors that are greater than the sum of their parts [1]. However, the term "swarm" has recently been appropriated in novel contexts, such as OpenAI's Swarm (OAS) framework [2], where the dynamics and mechanisms differ from their traditional counterparts. This paper explores the differences, examining how the principles that define classical swarm algorithms translate, or fail to translate, within large language model (LLM)-based systems such as OAS, which is selected as a representative framework for LLM-powered swarms in this paper.

Simulated Students offer a valuable methodological framework for evaluating pedagogical approaches and modelling diverse learner profiles, tasks which are otherwise challenging to undertake systematically in real-world settings. Recent research has increasingly focused on developing such simulated agents to capture a range of learning styles, cognitive development pathways, and social behaviours. Among contemporary simulation techniques, the integration of large language models (LLMs) into educational research has emerged as a particularly versatile and scalable paradigm. LLMs afford a high degree of linguistic realism and behavioural adaptability, enabling agents to approximate cognitive processes and engage in contextually appropriate pedagogical dialogues. This paper presents a thematic review of empirical and methodological studies utilising LLMs to simulate student behaviour across educational environments. We synthesise current evidence on the capacity of LLM-based agents to emulate learner archetypes, respond to instructional inputs, and interact within multi-agent classroom scenarios. Furthermore, we examine the implications of such systems for curriculum development, instructional evaluation, and teacher training. While LLMs surpass rule-based systems in natural language generation and situational flexibility, ongoing concerns persist regarding algorithmic bias, evaluation reliability, and alignment with educational objectives. The review identifies existing technological and methodological gaps and proposes future research directions for integrating generative AI into adaptive learning systems and instructional design.

Accessing legal help in Bangladesh is hard. People face high fees, complex legal language, a shortage of lawyers, and millions of unresolved court cases. Generative AI models like OpenAI GPT-4.1 Mini, Gemini 2.0 Flash, Meta Llama 3 70B, and DeepSeek R1 could potentially democratize legal assistance by providing quick and affordable legal advice. In this study, we collected 250 authentic legal questions from the Facebook group "Know Your Rights," where verified legal experts regularly provide authoritative answers. These questions were subsequently submitted to four four advanced AI models and responses were generated using a consistent, standardized prompt. A comprehensive dual evaluation framework was employed, in which a state-of-the-art LLM model served as a judge, assessing each AI-generated response across four critical dimensions: factual accuracy, legal appropriateness, completeness, and clarity. Following this, the same set of questions was evaluated by three licensed Bangladeshi legal professionals according to the same criteria. In addition, automated evaluation metrics, including BLEU scores, were applied to assess response similarity. Our findings reveal a complex landscape where AI models frequently generate high-quality, well-structured legal responses but also produce dangerous misinformation, including fabricated case citations, incorrect legal procedures, and potentially harmful advice. These results underscore the critical need for rigorous expert validation and comprehensive safeguards before AI systems can be safely deployed for legal consultation in Bangladesh.

The company is pursuing aggressive legal tactics against its opponents. On September 12, Jay Edelson received what he expected to be a standard legal document. Edelson is a lawyer representing the parents of Adam Raine; they are suing OpenAI, alleging that their 16-year-old son took his life at the encouragement of ChatGPT. OpenAI's lawyers had some inquiries for the opposing counsel, which is normal. For instance, they requested information about therapy Raine may have received, and Edelson complied.

If the US Has to Build Data Centers, Here's Where They Should Go A new analysis tries to calculate the coming environmental footprint of AI in the US and finds that the ideal sites for data centers aren't where they're being built. A data center for cryptocurrency mining, cloud services, and AI computing in Stutsman County, North Dakota.Video: halbergman/Getty Images Tech companies have invested so much money in building data centers in recent months, it's actively driving the US economy--and the AI race is showing no signs of slowing down. Meta chief Mark Zuckerberg told President Donald Trump last week that the company would spend $600 billion on US infrastructure--including data centers--by 2028, while OpenAI has committed already to spending $1.4 trillion. An extensive new analysis looks at the environmental footprint of data centers in the US to get a handle on what, exactly, the country might be facing as this buildout continues over the next few years--and where the US should be building data centers to avoid the most harmful environmental impacts. The study, published in the journal Nature Communications on Monday, uses a variety of data, including demand for AI chips and information on state electricity and water scarcity, to project the potential environmental impacts of future data centers through the end of the decade. The study models a number of different possible scenarios on how data centers could affect the US and the planet--and cautions that tech companies' net zero promises aren't likely to hold up against the energy and water needs of the massive facilities they're building.