Explore the vast range of titles available.

This paper introduces assignment flows for density matrices as state spaces for representation and analysis of data associated with vertices of an underlying weighted graph. Determining an assignment flow by geometric integration of the defining dynamical system causes an interaction of the non-commuting states across the graph, and the assignment of a pure (rank-one) state to each vertex after convergence. Adopting the Riemannian–Bogoliubov–Kubo–Mori metric from information geometry leads to closed-form local expressions that can be computed efficiently and implemented in a fine-grained parallel manner. Restriction to the submanifold of commuting density matrices recovers the assignment flows for categorical probability distributions, which merely assign labels from a finite set to each data point. As shown for these flows in our prior work, the novel class of quantum state assignment flows can also be characterized as Riemannian gradient flows with respect to a non-local, non-convex potential after proper reparameterization and under mild conditions on the underlying weight function. This weight function generates the parameters of the layers of a neural network corresponding to and generated by each step of the geometric integration scheme. Numerical results indicate and illustrate the potential of the novel approach for data representation and analysis, including the representation of correlations of data across the graph by entanglement and tensorization.

Purpose Originally, the coastal region of East Africa was largely covered with coastal forest. With the human settlement, the majority of these forests have been transformed into agricultural land for subsistence farming. Today, only small and geographically isolated forest remnants exist and form a mosaic of different habitat types, including natural, semi-natural and anthropogenic ones. The forest remnants may still represent valuable habitats for typical forest plant and animal species. Methods In this study, we surveyed hymenopteran diversity and community composition in different habitat types in southern Kenya. Hereby we considered a small remnant of East African coastal forest, adjoining orchards, shrublands, tree plantations, agricultural fields, and settlements. Hymenoptera represent a large variety of taxa and provide relevant ecosystem services such as pollination to the local people. Hymenoptera were collected with coloured pan traps, identified to family or genus level, and sorted to morphospecies. Habitat parameters such as dead wood, leaf litter, vegetation structure, and the availability of flowers has been assessed for each study site. Results In total, we found 419 Hymenoptera individuals of 153 morphospecies. The different habitat types showed distinct species communities of Hymenoptera. Increasing shrub layer and increasing herb cover had a negative effect on hymenopteran species richness and abundance. Flowers, trees, leave litter, and deadwood showed no significant effect on hymenopteran diversity. Conclusion Our results underline that small-scale habitat diversity lead to high overall diversity of Hymenoptera. Implications for insect conservation: The coastal forest remnant still holds a unique hymenopteran community with 20 Hymenoptera species exclusively found in this habitat type. It is thus of very high conservation value.

Reading comprehension (RC)—in contrast to information retrieval—requires integrating information and reasoning about events, entities, and their relations across a full document. Question answering is conventionally used to assess RC ability, in both artificial agents and children learning to read. However, existing RC datasets and tasks are dominated by questions that can be solved by selecting answers using superficial information (e.g., local context similarity or global term frequency); they thus fail to test for the essential integrative aspect of RC. To encourage progress on deeper comprehension of language, we present a new dataset and set of tasks in which the reader must answer questions about stories by reading entire books or movie scripts. These tasks are designed so that successfully answering their questions requires understanding the underlying narrative rather than relying on shallow pattern matching or salience. We show that although humans solve the tasks easily, standard RC models struggle on the tasks presented here. We provide an analysis of the dataset and the challenges it presents.

In this article, the author first provides a UK perspective on treaty shopping and treaty abuse. He considers judicial approaches to treaty shopping, examine anti-abuse measures in UK tax treaties, and look at the new domestic-law general anti-avoidance rule as it relates to treaties. Judicial reluctance to allow treaty shopping or other treaty abuse suggests that existing measures in treaties are sufficient and that a case has not been made for the introduction of general anti-abuse rules in this area. Broad domestic treaty overrides jeopardize the international legal order and rule of law. Second, he examines EU law by reference to the approach taken by the Court of Justice of the EU in relation to treaty shopping. The cases are essentially constitutional in nature because they test the validity of tax treaty provisions against the fundamental freedoms in the treaties constituting the EU as taxpayers challenge the jurisdiction of member states to conclude treaties containing limitation-on-benefits provisions.

Test-time scaling (TTS) techniques can improve the performance of large language models (LLMs) at the expense of additional computation and latency. While TTS has proven effective in formal domains such as mathematics and programming, its value in argumentative domains such as law remains underexplored. We present an empirical study of verifier-based TTS methods for legal multiple-choice QA (MCQA) across five benchmarks. Using a family of 7 reward models, we evaluate both outcome-level (Best-of-\\(N\\)) and process-level (tree search) verification under realistic low-\\(N\\) budgets. Our analysis systematically investigates how verifier utility is affected by key properties such as domain specialization, model size, and supervision type (process-supervised PRMs vs. outcome-only ORMs), even when applied across different roles.

Diese Masterarbeit handelt von Transzendenzmomenten angesichts des Anderen und nimmt damit Bezug auf einen der einflussreichsten Philosophen der Gegenwart, Emmanuel Levinas. Philosophiegeschichtlich bildet der linguistic turn den Kontext dieses Diskurses. So wird der Wandel im Denken, der mit dem linguistic turn einhergeht, anhand verschiedener philosophischer und theologischer Essays reflektiert und auf das Problem der Gewalt im Prozess des Erkennens hin zugespitzt. In Diskussion mit den Schriften Dietrich Bonhoeffers leistet diese Arbeit hinfort einen Beitrag zum systematisch-theologischen Diskurs über Gottesoffenbarung in zwischenmenschlichen Beziehungen und über Ethik. In Auseinandersetzung mit Levinas und Bonhoeffer baut diese Arbeit eine Brücke zwischen postmodernem, dekonstruktivistischem Denken und der fortwährenden theologischen Aufgabe, Gottes Sein mittels menschlicher Sprache Ausdruck zu verleihen.

We present Sparse Interpolated Mixture-of-Experts (SIMoE) instruction-tuning, an end-to-end algorithm designed to fine-tune a dense pre-trained Large Language Model (LLM) into a MoE-style model that possesses capabilities in multiple specialized domains. During instruction-tuning, SIMoE automatically identifies multiple specialized experts under a specified sparsity constraint, with each expert representing a structurally sparse subset of the seed LLM's parameters that correspond to domain-specific knowledge within the data. SIMoE simultaneously learns an input-dependent expert merging strategy via a router network, leveraging rich cross-expert knowledge for superior downstream generalization that surpasses existing baselines. Empirically, SIMoE consistently achieves state-of-the-art performance on common instruction-tuning benchmarks while maintaining an optimal performance-compute trade-off compared to all baselines.

Large language model (LLM) post-training enhances latent skills, unlocks value alignment, improves performance, and enables domain adaptation. Unfortunately, post-training is known to induce forgetting, especially in the ubiquitous use-case of leveraging third-party pre-trained models, which is typically understood as a loss of parametric or factual knowledge. We argue that this accuracy-centric view is insufficient for modern foundation models and instead define forgetting as systematic model drift that degrades behavior and user experience. In this context, we introduce \\textbf{CapTrack}, a capability-centric framework for analyzing forgetting in LLMs that combines a behavioral taxonomy with an evaluation suite built on established benchmarks and targeted adaptations. Using CapTrack, we conduct a large-scale empirical study across post-training algorithms, domains, and model families, including models up to 80B parameters. We find that forgetting extends beyond parametric knowledge, with pronounced drift in robustness and default behaviors. Instruction fine-tuning induces the strongest relative drift, while preference optimization is more conservative and can partially recover lost capabilities. Differences across model families persist, and no universal mitigation emerges.

LLM post-training has primarily relied on large text corpora and human feedback, without capturing the structure of domain knowledge. This has caused models to struggle dealing with complex reasoning tasks, especially for high-stakes professional domains. In Law, reasoning requires deep understanding of the relations between various legal concepts, a key component missing in current LLM post-training. In this paper, we propose a knowledge graph (KG)-assisted approach for enhancing LLMs' reasoning capability in Legal that is generalizable to other high-stakes domains. We model key legal concepts by following the \\textbf{IRAC} (Issue, Rule, Analysis and Conclusion) framework, and construct a KG with 12K legal cases. We then produce training data using our IRAC KG, and conduct both Supervised Fine-Tuning (SFT) and Direct Preference Optimization (DPO) with three state-of-the-art (SOTA) LLMs (30B, 49B and 70B), varying architecture and base model family. Our post-trained models obtained better average performance on 4/5 diverse legal benchmarks (14 tasks) than baselines. In particular, our 70B DPO model achieved the best score on 4/6 reasoning tasks, among baselines and a 141B SOTA legal LLM, demonstrating the effectiveness of our KG for enhancing LLMs' legal reasoning capability.