arXiv:2607.09990v1 Announce Type: cross
Abstract: Economic power in international trade is the capacity of one country to impose loss on another by withdrawing from a trading relationship. This paper measures it. A model of the short run represents each trade restriction as a pattern of barred entries in the world matrix of input shares and maps it into a vector of losses by country and sector. The asymmetry between the two countries' losses under the same severance is the measure of power: a gap in substitution, since a buyer's dependence turns on how easily it finds another source and a seller's on how easily it finds another market. When a relationship is barred, buyers lean on alternative suppliers and barred suppliers on alternative buyers already present in the benchmark network, under the ceiling that no producer exceeds its pre-shock scale. The reallocation is a RAS balancing of the disrupted matrix that lets \emph{both sides} adjust together. Across 9,480 counterfactual severances on the 2022 world input--output network, mutual trade dependence is anything but mutual. The average bilateral asymmetry is 0.6 on a scale that runs from balance at zero to complete lopsidedness at one. The United States holds the favorable side in all its relationships, China in all but one. The same tilt runs far down the hierarchy: a severance with Russia would cost Belarus more than a tenth of its economic activity, but Russia only half of one percent. The asymmetry bears only a weak relation to bilateral trade imbalance but closely tracks whether a country sits at the core or the periphery. Power is a property of network position, not deficits.
Science Journals
arXiv:2607.10923v1 Announce Type: new
Abstract: Large language models exhibit remarkable emergent behaviors, yet the physical mechanism governing their collective dynamics remains poorly understood. Cognitive Field Theory predicts that learning reorganizes the time-scale density of states (TDOS) through the infrared accumulation of slow relaxation modes, thereby enhancing the memory self-energy, reducing the cognitive forgetting gap, and strengthening the collective susceptibility. Using publicly available Pythia language models, we extract relaxation spectra directly from Transformer layer Jacobians throughout training, network depth, and model scale, allowing the TDOS, memory self-energy, forgetting gap, memory kernel, and infrared critical exponent to be measured quantitatively. The measurements reveal progressive infrared accumulation of slow relaxation modes, producing an approximately flat infrared TDOS with \( \rho(\lambda)\sim\lambda^{-0.1} \) and scale-free memory kernels \( K(t)\sim t^{-1}. \) The memory self-energy exhibits a pronounced transient maximum during early optimization before relaxing toward a metastable near-critical regime, corresponding to the smallest cognitive forgetting gap and the largest collective susceptibility predicted by Cognitive Field Theory. These observations provide quantitative experimental evidence that Transformer dynamics are governed by infrared collective organization. The reproducibility of the same dynamical behavior across training, network depth, and model scales suggests that infrared slow-mode organization represents a universal collective principle of Transformer dynamics.
arXiv:2607.10927v1 Announce Type: new
Abstract: Astroparticle physics emerged during the late twentieth century as a new interdisciplinary field at the intersection of particle physics, astrophysics, and cosmology. This article examines the historical mechanisms through which it took shape as a distinct scientific field, focusing on the progressive interaction of previously separate experimental cultures, research traditions, and scientific communities. It argues that astroparticle physics emerged not through the simple convergence of established disciplines, but through the gradual reorganization of high-energy research around scientific problems that transcended traditional disciplinary boundaries. Three broad historical trajectories shaped this process: the transformation of postwar cosmic-ray physics; the emergence of a high-energy Universe through relativistic astrophysics and new astronomical windows; and the growing interaction between particle physics and cosmology. Their progressive interaction transformed naturally occurring particles into astrophysical messengers and the Universe itself into a laboratory for investigating fundamental physics beyond the reach of terrestrial accelerators. At the same time, new experimental practices, international collaborations, scientific forums, and institutional structures progressively gave organizational form to this emerging research landscape. More broadly, the emergence of astroparticle physics illustrates how new interdisciplinary fields develop through the co-evolution of scientific questions, experimental cultures, research communities, and institutions.
arXiv:2607.11475v1 Announce Type: new
Abstract: Safety alignment in large language models can be fragile under fine-tuning, as even benign task adaptation may increase harmful compliance. Existing defenses mainly follow two directions: they either intervene during or after fine-tuning through retraining or weight modification, which can be costly and may hurt task performance, or they use model-agnostic safety classifiers, which may miss failures specific to a given fine-tuned checkpoint. These limitations motivate a post hoc, model-specific, and non-invasive approach to safety restoration. To meet these requirements, we propose HyperSafe, a framework that restores safety behavior by generating a model-specific Safe Side Network (SSN) for each fine-tuned checkpoint. HyperSafe uses layer-wise activation fingerprints to capture how fine-tuning changes the model's inner representations. With a small set of given calibration prompts, the hypernetwork maps these fingerprints to the parameters of the \ssn{} in a single forward pass. The generated \ssn{} runs alongside the frozen fine-tuned model and performs prompt-level safety classification: harmful prompts are routed to refusal, while safe prompts are answered by the original fine-tuned model. Thus, HyperSafe requires no gradient updates, no safety data at deployment time, and no modification to the deployed model weights. We evaluate HyperSafe on two model families, Qwen2-7B and LLaMA-3-8B, across multiple safety benchmarks. HyperSafe reduces harmful response rates from 19-31% to below 1% on every held-out checkpoint, while keeping downstream task accuracy within 1% of the fine-tuned baseline on average. Code is available at https://github.com/nokronim/project-safety-remedy.
arXiv:2607.11739v1 Announce Type: new
Abstract: Real-time navigation in cluttered and dynamic environments requires collision-free and dynamically feasible motion under limited perception. However, feasible navigation behaviors are inherently multimodal because multiple paths may exist around obstacles. In this paper, we formulate navigation as learning a transferable goal-conditioned stochastic path prior that models a reusable distribution over goal-aligned geometry-consistent local paths conditioned on local observations. This formulation enables structured sampling of navigation candidates, allowing multiple feasible paths to be explored through sampling without relying on robot-specific motion constraints. To this end, we introduce a goal-aligned canonical state representation that removes in-plane rotational ambiguity and normalizes local geometry with respect to the goal, enabling rotation-invariant path distribution learning. We further develop a structured prior learning framework that parameterizes local paths using a geometry-aware polar action manifold and incorporates risk-sensitive utility shaping with multi-goal distributional rollouts for stable and safety-aware planning. Extensive experiments in dense static environments and dynamic pedestrian scenarios demonstrate that the proposed method achieves consistently high success rates with competitive efficiency while enabling cross-platform transfer of a single path prior learned on differential-drive robots to quadruped platforms without retraining.
arXiv:2602.22395v2 Announce Type: cross
Abstract: Since 2016, the term "misinformation" has become associated with a scientific paradigm that studies, at its core, people making, reading, and sharing false statements, usually on social media, and often warning of the harm to society resulting from the sum of many such events. By tracking the term through the academic literature, with special focus on the years 2011--2023, we connect the post-2016 paradigm with a strand of research dating to the Satanic panic of the 1980s. We argue that post-2016 misinformation research owes more to this intellectual lineage than is generally acknowledged, and we discuss the theoretical and practical implications of this connection. We conclude by drawing parallels between the Satanic panic and 2026, and, similarly, between misinformation research then and now.
arXiv:2607.11492v1 Announce Type: new
Abstract: In this paper, we investigate preprocessing techniques aimed at improving the efficiency of accessing models of propositional formulas represented in conjunctive normal form (CNF). We focus on three fundamental tasks: uniform sampling, direct model access, and model enumeration. Our analysis reveals that most state-of-the-art preprocessors, when they do not preserve formula equivalence, are generally unsuitable for these tasks. In contrast, we demonstrate that preprocessors which preserve model counts can be effectively leveraged, provided relevant preprocessing information is maintained. To validate our approach, we perform extensive experiments on a diverse suite of benchmarks from multiple domains. The experimental results show that our preprocessing methods are both efficient and robust, yielding significant performance improvements for model access queries when CNF formulas are compiled into d-DNNF representations.
arXiv:2607.11817v1 Announce Type: new
Abstract: Since the foundational studies in the late nineteenth century, fluid turbulence has stood as a profound, unsolved challenge in classical physics. Much of this enduring difficulty stems from non-equilibrium turbulence, where the lack of a unifying physical framework for macroscopic coherent structures has hampered predictive flow modeling. Here, we establish a foundational bridge between non-equilibrium statistical physics and turbulent coherent structures through the renormalized Navier-Stokes equations. We demonstrate that all forms of turbulent coherence are fundamentally universal thermodynamic responses mandated by macroscopic energy throughput imbalances. Depending on topological access to bifurcations, these formations manifest either as transient adjustments (analogous to Kubo's near-equilibrium fluctuations) or as autonomous, transformative states (mirroring Prigogine's far-from-equilibrium dissipative structures). By introducing a computable, effective thermodynamic order parameter ($\Pi$), this paradigm establishes a rigorous foundation for non-equilibrium theory, enabling unequivocal identification of necessary flow resolution in continuously driven, dissipative continuum systems.
arXiv:2607.10233v1 Announce Type: new
Abstract: Melody skeleton extraction aims to derive a shorter melody that preserves structural notes while removing ornaments. Prior methods rely on hand-crafted reduction rules or note-wise salience classifiers trained with heuristically or procedurally generated pseudo-labels. Such supervision can inherit generator bias and does not explicitly optimize a coherent reduced melody. We introduce MeloBottleneck, a self-supervised framework that represents a skeleton as a length-controlled, order-preserving latent subsequence. A hard-bottleneck extractor selects note events, a rhythmic-closure operator produces a self-consistent skeleton, and a re-ornamentation decoder reconstructs the input melody. Training combines reconstruction, a frozen autoregressive melody prior, ornament-invariant consistency across procedurally ornamented views, and ornament exclusion. We evaluate three regimes: synthetic out-of-distribution ornament-to-skeleton, TAVERN variation-to-theme, and Jiugong ornamented-to-gongche. A matched pseudo-label classifier excels on the synthetic benchmark, while MeloBottleneck transfers better, achieving competitive selection quality on TAVERN and Jiugong. Skeletonized melodies also improve BM25-based fragment retrieval, boosting Recall@K and MRR while reducing query time. Overall, the results suggest that learning skeletons as latent subsequences yields more robust transfer than pseudo-label imitation.
arXiv:2607.09755v1 Announce Type: new
Abstract: Urban rail fare systems may be non-additive: the fare of a single paid journey from an origin to a destination can differ from the sum of fares over multiple legally separated journey legs. This paper presents LegalFarePlan, a fare-transparent route-planning framework that models legal exit-and-reentry operations as explicit, auditable constraints. Given a transit network, fare function, transfer rules, station-level exit/re-entry costs, an extra-time budget, and a split limit, the planner computes explainable route plans over paid journey segments. The artifact implements Dijkstra shortest-time and direct route-planner baselines, a greedy split heuristic, bounded exact label-setting, and Pareto-frontier search. Evaluation uses controlled synthetic data and a 57-station semi-synthetic benchmark with 360 OD pairs. On the semi-synthetic benchmark, bounded exact search identifies positive modeled fare reductions for 71.11% of OD pairs, with mean reduction 3.78 and maximum reduction 9.0 synthetic fare units under a 45-minute extra-time budget. These results demonstrate method behavior and reproducibility; they are not empirical conclusions about MTR or any transit operator.
arXiv:2607.09758v1 Announce Type: new
Abstract: We consider a model of chemical kinetics for which the derivation of equations does not rely on the law of mass action, but is rather based on such principles as the joint probability and the geometric probability. For this model a generalization is constructed for the case of reaction-diffusion systems in heterogeneous medium with respect to the convective and diffusive transfer of heat. The construction of this generalization is carried out by an alternative methodology which is based fully on a systems of ordinary differential equations, without a transition to the partial derivatives. The description of this new method is a bit similar to the finite volume method, except that it uses statistical simplifying positions and geometric probability to describe the diffusion processes. Such approach allows us to greatly simplify the numerical implementation of the resulting model, as well as to simplify the quantitative analysis of it with dynamical systems theory. Moreover, the efficiency of the parallel implementation of the numerical method is increased for the resulting model. In addition, we will consider an application of this model for the description of some example reaction with quasi-periodic regime, as well as consider an algorithm for the transition from standard models with dimensional kinetic constants to its formalism.
arXiv:2607.11138v1 Announce Type: new
Abstract: The rapid expansion of capabilities in Large Language Model (LLM) agents has exposed a critical architectural bottleneck: when agents are given access to a flat, monolithic registry of tools, the model must evaluate hundreds or thousands of options simultaneously. This leads to decision-space explosion, context window saturation, and degraded routing accuracy. To address these limitations, this paper presents a hierarchical, skill-based architecture for agentic orchestration. Capabilities are organized as a rooted tree where internal nodes make routing decisions and leaf nodes execute deterministic tasks. The runtime enforces a single-step execution loop governed by a Last-In-First-Out (LIFO) stack, giving the agent a form of memory akin to a Pushdown Automaton, therefore enabling it to track nested execution contexts and resume deterministically from any depth. Capability discovery follows a manifest-driven, lazy-loading protocol: only the immediate children of the active node are loaded, so memory and prompt costs scale with the explored path rather than the global registry. By replacing global memory with localized stack frames, the architecture prevents outputs from one execution branch from leaking into another, establishing the isolation guarantees required for deployment in regulated enterprise environments. We also discuss UPI Help, an AI-powered digital payments support product, as a motivating production deployment context. We provide a mathematical formalization of the orchestration state, detailed algorithmic analysis of the execution loop, and controlled benchmarks comparing flat and hierarchical routing under increasing tool catalogs, multi-step workflow pressure, and visible schema-token exposure per LLM call.
arXiv:2509.05211v2 Announce Type: replace
Abstract: We introduce a new technique for proving bounds on the Kolmogorov complexity of geometric objects in Euclidean space, such as points and lines. We apply this technique to prove two theorems on algorithmic information theory, both of which have consequences for well-known problems in geometric measure theory. First, we show that for any point $x$ in the plane and any other point $y$ sufficiently independent of $x$, the distance between $x$ and $y$ retains at least half the complexity of the original point $x$. By the point-to-set principle of J. Lutz and N. Lutz, this yields an improved lower bound on the Hausdorff dimension of pinned distance sets, a topic closely related to Falconer's distance set conjecture. Second, we prove an analogous result for orthogonal projections: for any point $x$ in the plane and any line through the origin which is sufficiently independent of $x$, the projection of $x$ onto that line retains at least half the complexity of $x$. As a consequence, we obtain a generalization of a theorem of Bourgain on exceptional sets for orthogonal projections.
arXiv:2509.24148v3 Announce Type: replace
Abstract: Test-Driven Development (TDD) is a widely adopted practice that requires developers to create and execute tests alongside implementation. With recent advances in Large Language Models (LLMs), developers can shift from manually writing the code to defining tests as executable specifications and delegating code synthesis to AI agents. However, enabling repository-level TDD under developer-written tests is challenging, requiring: (1) specification enhancement: identifying a concise yet representative test subset from large suites with rich task semantics; (2) retrieval augmentation: using tests to guide reasoning and context retrieval; and (3) test-driven refinement: interpreting test feedback for iterative improvement. We propose TENET, an agentic framework for repository-level code generation under the TDD paradigm. TENET includes: (1) a test harness mechanism that selects a concise test suite to maximize diversity of the target usage scenarios; (2) a tailored agent toolset for efficient retrieval and debugging; and (3) a reflection-based refinement workflow that iteratively analyzes failures and updates implementations. TENET consistently outperforms the strongest baselines across backbones, achieving 69.08% and 81.77% Pass@1 on RepoCod and RepoEval with Claude Sonnet 4, improving by 9.49 and 2.17 percentage points, respectively. Additionally, we present the first systematic study of how test suite characteristics influence LLM agent performance in TDD settings.
arXiv:2607.11306v1 Announce Type: cross
Abstract: This paper addresses the optimal control of epidemic dynamics under conflicting socio-economic objectives. We propose an economic Model Predictive Control (MPC) framework, applied to an extended SEIR-V (Susceptible-Exposed-Infected-Recovered-Vaccinated) compartmental model to govern the spread of an infectious disease while minimizing economic disruption. The control problem is formulated as a constrained nonlinear optimization problem, in which the controller dynamically adjusts social interaction levels (transmission rate beta) and vaccination efforts to minimize a composite cost function that penalizes fatalities, healthcare capacity violations, and economic losses. We conduct a rigorous sensitivity analysis of the prediction horizon N, demonstrating that the closed loop is robust to the horizon choice and that N = 35 days minimizes the realized cost. Furthermore, both the closed-loop solution and an open-loop turnpike analysis across diverse initial conditions reveal that the celebrated "Hammer and Dance" mitigation strategy emerges naturally as the mathematical optimum: the optimal trajectories anchor to a unique suppression turnpike (maximum lockdown) to drive hospitalizations toward the disease-free equilibrium before progressively reopening the economy. Through a turnpike-based argument we establish practical asymptotic stability of the optimal operating point, providing a mathematically grounded decision-support tool for pandemic policy.
arXiv:2607.11344v1 Announce Type: cross
Abstract: In this work, we consider the identification and control of nonlinear systems with finite action spaces. The unknown dynamics are estimated from finite samples with Koopman operator regression in a reproducing kernel Hilbert space, yielding a linear switching predictive model, the switches governed by the value of the control variable. In order to perform control in closed-loop, the learned dynamics are employed in an infinite-horizon optimal control problem with time-varying stage cost, which is solved by means of model predictive control. In a theoretical analysis, we derive learning rates for the Koopman dynamics approximation. We further quantify, under suitable assumptions, the sub-optimality of the model predictive control strategy, both in the case of exact Koopman dynamics, and in the case of learned ones. Numerical simulations on the Duffing oscillator complement our theoretical findings.
arXiv:2607.10994v1 Announce Type: new
Abstract: Zero-dimensional reduced-order models (0D ROMs) are central to multi-dimensional design workflows for high-end complex equipment. However, the planning process currently relies on manual expertise, limiting topological exploration and prolonging iterations. Even traditional optimization methods such as Genetic Algorithms (GA) are typically confined to local parameter tuning. Although Large Language Model (LLM) agents have shown promise in exploring large sample spaces, and frameworks such as Chain of Thought (CoT) and Reason and Act (ReAct) improve reasoning reliability, while Retrieval-Augmented Generation (RAG) overcomes domain knowledge barriers, a single agent still falls short for the long-horizon and highly coupled nature of complex 0D ROM planning. This paper proposes the Zero-dimensional reduced-order model CO-Planning framework (Z-COPA), a multi-agent architecture featuring a Symbolic Action Graph Engine (SAGE) and a MILP-Guided Navigation (MGN) optimizer. Its core innovation is a dedicated graph representation method that accurately encodes the 0D flow network topology, converting the empirical planning process into a rigorous graph structure optimization problem. We validate the forward and inverse design capabilities and generalization performance of Z-COPA on two real aircraft engine secondary-air systems, two IEEE power-distribution reconfiguration benchmarks, and two water-distribution network benchmarks. The results show superior task completion quality, obtaining the best performance in both forward and reverse design of air systems. Z-COPA disrupts the traditional 0D model planning paradigm, providing a new technical approach for exploring broader topological space and achieving highly automated, globally optimal air system architectures.
Decoupling Constraints from Two Directions for Evolutionary Constrained Multi-objective Optimization
arXiv:2512.23945v2 Announce Type: replace
Abstract: Real-world constrained multi-objective optimization problems (CMOPs) commonly involve multiple constraints, and understanding and exploiting their coupling relationships is crucial for efficient optimization. Recent constraint-decoupling methods handle individual constraints separately, but they generally search only in the evolutionary direction to approximate single-constraint Pareto fronts (SCPFs). In this study, we show that part or all of the constrained Pareto front (CPF) may be unrelated to any SCPF and instead be shaped by the boundaries of infeasible regions. We refer to such a portion as the independent CPF (ICPF) and introduce the reverse CPF (RCPF) to characterize its associated informative infeasible boundaries. Based on these observations, we propose a bidirectional constraint-decoupling coevolutionary algorithm named DCF2D. DCF2D dynamically identifies the constraints obstructing the main population and activates constraint-specific auxiliary populations. These populations adaptively search in the evolutionary direction for the corresponding SCPFs or in the reverse evolutionary direction for the corresponding RCPFs. Its three-stage framework integrates unconstrained global exploration, event-driven bidirectional coevolution, and final convergence refinement. Experiments on 87 benchmark instances from seven test suites and 28 real-world engineering CMOPs demonstrate that DCF2D achieves the best overall performance among nine algorithms. Code available at: https://github.com/RuiqingS/DCF2D.
arXiv:2601.01406v3 Announce Type: replace
Abstract: Face super-resolution aims to recover high-quality facial images from severely degraded low-resolution inputs, but remains challenging due to the loss of fine structural details and identity-specific features. This work introduces SwinIFS, a landmark-guided super-resolution framework that integrates structural priors with hierarchical attention mechanisms to achieve identity-preserving reconstruction at both moderate and extreme upscaling factors. The method incorporates dense Gaussian heatmaps of key facial landmarks into the input representation, enabling the network to focus on semantically important facial regions from the earliest stages of processing. A compact Swin Transformer backbone is employed to capture long-range contextual information while preserving local geometry, allowing the model to restore subtle facial textures and maintain global structural consistency. Extensive experiments on the CelebA benchmark demonstrate that SwinIFS achieves superior perceptual quality, sharper reconstructions, and improved identity retention; it consistently produces more photorealistic results and exhibits strong performance even under $8\times$ magnification, where most methods fail to recover meaningful structure. SwinIFS also provides an advantageous balance between reconstruction accuracy and computational efficiency, making it suitable for real-world applications in facial enhancement, surveillance, and digital restoration. Our code, model weights, and results are available at https://github.com/Habiba123-stack/SwinIFS.
arXiv:2601.05636v2 Announce Type: replace
Abstract: We study deletion-correcting codes in the space of length-$n$ multisets over a $q$-ary alphabet. We present an explicit cyclic Sidon-type construction for arbitrary alphabet size $q$ and deletion radius $t$, defined by a single congruence modulo $t(t+1)^{q-2}+1$. The construction has redundancy at most $\log_q(t(t+1)^{q-2}+1)$ and admits linear-time online decoding for fixed $q$ and $t$ after finite preprocessing. We prove that its syndrome classes are asymptotically balanced and compare several general upper bounds. For a single deletion, we show that the natural sum-modulo construction is asymptotically optimal for every fixed $q$. We also obtain exact results for $q=3$ and $q=4$, including uniqueness results for optimal codes in the relevant parameter ranges, and formulate conjectures for prime alphabets.
arXiv:2601.06362v2 Announce Type: replace
Abstract: Personalized large language models are often expected to follow explicit style instructions, yet we find that such instructions can undermine the user-specific characteristics that personalization methods aim to preserve. We call this failure mode personalization collapse: explicit style control can conflict with implicit user preferences. To address this challenge, we propose PsPLUG, a lightweight plug-in that learns a user-specific residual after accounting for the requested style. PsPLUG also allows us to tune personalization strength at inference time. Our experiments show that explicit style instructions can diminish personalization in existing methods, whereas PsPLUG better preserves user preferences while providing precise control over the balance between personalization and style adherence.
arXiv:2601.07155v3 Announce Type: replace
Abstract: Knowledge distillation (KD) is a widely adopted technique for transferring knowledge from large language models to smaller student models; however, conventional supervised KD often suffers from a distribution mismatch between training and inference. While on-policy KD approaches attempt to mitigate this issue by learning directly from student-generated outputs, they frequently encounter training instabilities because the distributional gap between the novice student and the expert teacher is often too wide to bridge directly. These challenges manifest as pathological gradients in forward KL objectives or diversity collapse in reverse KL regimes. To address these limitations, we propose Veto, an objective-level reformulation that constructs a geometric bridge in the logit space. Unlike prior methods that mix data samples, Veto creates an intermediate target distribution that promotes alignment between the teacher and the student. By introducing a tunable parameter beta, Veto serves as an Adaptive Gradient Veto that stabilizes optimization by suppressing harmful gradients on low-confidence tokens, while simultaneously acting as a Decisiveness Knob to balance reward-driven performance with output diversity. Extensive experiments across various reasoning and generation tasks demonstrate that Veto consistently outperforms supervised fine-tuning and existing on-policy baselines.
arXiv:2607.10794v1 Announce Type: new
Abstract: Model stealing attacks have recently been introduced, enabling the extraction of precise information from black-box commercial language models. In this work, we propose defense methods against a recent attack of \cite{carlini2024stealing} and extensions for extracting the hidden layer dimension of production language models. Our methods are inspired by watermarking techniques that perturb the logits layer of these models to prevent such attacks. We provide empirical experiments demonstrating the effectiveness of the proposed defense versus model quality degradation across various configurations, and propose an effective defense against such attacks while preserving model utility.
arXiv:2607.11784v1 Announce Type: new
Abstract: Equilibrium sheared $\vec{E}\times \vec{B}$ flow, a standard cure for plasma turbulence, can backfire. In gyrokinetic simulations of a newly identified regime, imposed shear comparable to the intrinsic zonal shear destroys the self-generated zonal flows regulating the turbulence: transport rises sharply before stronger shear quenches it. A reduced fluid model traces this to the incompatibility of imposed and zonal shear layers. Simulations of spherical tokamak discharges place the inferred rotation shear at, or just below, the threshold of the sharp transport increase, implying that the toroidal rotation may be limited mainly by the heat, not momentum, injection.
arXiv:2607.11760v1 Announce Type: new
Abstract: A theoretical understanding of Transformers is crucial to better understand the capacities and limitations of large language models (LLMs). There is much work analyzing the expressivity of attention-based models. By proposing handcrafted weights or using computational complexity arguments, a large amount of past theoretical works have sought to characterize which tasks are and which are not in the hypothesis class of Transformer models. However, little work investigates the learnability of such solutions. In this work, we make progress towards this goal. Inspired by recent loss landscape analysis work, we propose preliminary sample complexity bounds for learning C-RASP constructions with Transformers.