Can an AI autonomously design mechanisms for computer systems on par with the creativity and reasoning of human experts? We present Glia, an AI architecture for networked systems design that uses large language models (LLMs) in a human-inspired, multi-agent workflow. Each agent specializes in reasoning, experimentation, and analysis, collaborating through an evaluation framework that grounds abstract reasoning in empirical feedback. Unlike prior ML-for-systems methods that optimize black-box policies, Glia generates interpretable designs and exposes its reasoning process. When applied to a distributed GPU cluster for LLM inference, it produces new algorithms for request routing, scheduling, and auto-scaling that perform at human-expert levels in significantly less time, while yielding novel insights into workload behavior. Our results suggest that by combining reasoning LLMs with structured experimentation, an AI can produce creative and understandable designs for complex systems problems.

Connected health is a multidisciplinary approach focused on health management, prioritizing pa-tient needs in the creation of tools, services, and treatments. This paradigm ensures proactive and efficient care by facilitating the timely exchange of accurate patient information among all stake-holders in the care continuum. The rise of digital technologies and process innovations promises to enhance connected health by integrating various healthcare data sources. This integration aims to personalize care, predict health outcomes, and streamline patient management, though challeng-es remain, particularly in data architecture, application interoperability, and security. Data analytics can provide critical insights for informed decision-making and health co-creation, but solutions must prioritize end-users, including patients and healthcare professionals. This perspective was explored through an agile System Development Lifecycle in an EU-funded project aimed at developing an integrated AI-generated solution for managing cancer patients undergoing immunotherapy. This paper contributes with a collaborative digital framework integrating stakeholders across the care continuum, leveraging federated big data analytics and artificial intelligence for improved decision-making while ensuring privacy. Analytical capabilities, such as treatment recommendations and adverse event predictions, were validated using real-life data, achieving 70%-90% accuracy in a pilot study with the medical partners, demonstrating the framework's effectiveness.

However, although these techniques have been widely adopted in single-task scenarios, research is scarce in multi-task scenarios.

Neural Information Processing Systems http://nips.cc/

For instance, in clinical diagnosis, fully-autonomous machine behavior is often beyond ethical affordances, thus real-world decision support is often limited to monitoring and forecasting.

Standard training metrics like loss fail to explain the emergence of complex capabilities in large language models. We take a spectral approach to investigate the geometry of learned representations across pretraining and post-training, measuring effective rank (RankMe) and eigenspectrum decay ($α$-ReQ). With OLMo (1B-7B) and Pythia (160M-12B) models, we uncover a consistent non-monotonic sequence of three geometric phases during autoregressive pretraining. The initial "warmup" phase exhibits rapid representational collapse. This is followed by an "entropy-seeking" phase, where the manifold's dimensionality expands substantially, coinciding with peak n-gram memorization. Subsequently, a "compression-seeking" phase imposes anisotropic consolidation, selectively preserving variance along dominant eigendirections while contracting others, a transition marked with significant improvement in downstream task performance. We show these phases can emerge from a fundamental interplay of cross-entropy optimization under skewed token frequencies and representational bottlenecks ($d \ll |V|$). Post-training further transforms geometry: SFT and DPO drive "entropy-seeking" dynamics to integrate specific instructional or preferential data, improving in-distribution performance while degrading out-of-distribution robustness. Conversely, RLVR induces "compression-seeking", enhancing reward alignment but reducing generation diversity.

Large Language Models (LLMs) struggle with complex Text-to-SQL queries that demand both sophisticated mathematical reasoning and intricate schema navigation. Existing methods often tackle these challenges in isolation, creating a fractured reasoning process that compromises logical and structural correctness. To resolve this, we introduce SteinerSQL, a framework that unifies these dual challenges into a single, graph-centric optimization problem. SteinerSQL operates in three stages: mathematical decomposition to identify required tables (terminals), optimal reasoning scaffold construction via a Steiner tree problem, and multi-level validation to ensure correctness. On the challenging LogicCat and Spider2.0-Lite benchmarks, SteinerSQL establishes a new state-of-the-art with 36.10% and 40.04% execution accuracy, respectively, using Gemini-2.5-Pro. Beyond accuracy, SteinerSQL presents a new, unified paradigm for Text-to-SQL, paving the way for more robust and principled solutions to complex reasoning tasks.

Patent examiners and inventors face significant pressure to verify the originality and non-obviousness of inventions, and the intricate nature of patent data intensifies the challenges of patent retrieval. Therefore, there is a pressing need to devise cutting-edge retrieval strategies that can reliably achieve the desired recall. This study introduces FullRecall, a novel patent retrieval approach that effectively manages the complexity of patent data while maintaining the reliability of relevance matching and maximising recall. It leverages IPC-guided knowledge to generate informative phrases, which are processed to extract key information in the form of noun phrases characterising the query patent under observation. From these, the top k keyphrases are selected to construct a query for retrieving a focused subset of the dataset. This initial retrieval step achieves complete recall, successfully capturing all relevant documents. To further refine the results, a ranking scheme is applied to the retrieved subset, reducing its size while maintaining 100% recall. This multi-phase process demonstrates an effective strategy for balancing precision and recall in patent retrieval tasks. Comprehensive experiments were conducted, and the results were compared with baseline studies, namely HRR2 [1] and ReQ-ReC [2]. The proposed approach yielded superior results, achieving 100% recall in all five test cases. However, HRR2[1] recall values across the five test cases were 10%, 25%, 33.3%, 0%, and 14.29%, while ReQ-ReC [2] showed 50% for the first test case, 25% for the second test case, and 0% for the third, fourth, and fifth test cases. The 100% recall ensures that no relevant prior art is overlooked, thereby strengthening the patent pre-filing and examination processes, hence reducing potential legal risks.

With the growing practical interest in vision-based tasks for autonomous systems, the need for efficient and complex methods becomes increasingly larger. In the rush to develop new methods with the aim to outperform the current state of the art, an analysis of the underlying theory is often neglected and simply replaced with empirical evaluations in simulated or real-world experiments. While such methods might yield favorable performance in practice, they are often less well understood, which prevents them from being applied in safety-critical systems. The goal of this work is to design an algorithm for the Next Best View (NBV) problem in the context of active object reconstruction, for which we can provide qualitative performance guarantees with respect to true optimality. To the best of our knowledge, no previous work in this field addresses such an analysis for their proposed methods. Based on existing work on Gaussian process optimization, we rigorously derive sublinear bounds for the cumulative regret of our algorithm, which guarantees near-optimality. Complementing this, we evaluate the performance of our algorithm empirically within our simulation framework. We further provide additional insights through an extensive study of potential objective functions and analyze the differences to the results of related work.