Analysis of PDEs
Local Bifurcations from Single-Mode Traveling Waves in the Filamentation Equation
Yuri Cacchiò
We study local bifurcations from the single-mode traveling wave family of the filamentation equation on the
torus, within the real positive-frequency Sobolev space $X^s$. A
Fourier null structure removes the apparent derivative loss and makes the traveling wave map real analytic for $s>3/2$. Its linearization splits into finitely many coupled two-mode blocks, a carrier block, and a diagonal high-frequency tail. For negative temporal frequency, we determine the critical values generated by both the finite coupled blocks and the tail. At every non resonant finite-block critical value, Lyapunov-Schmidt reduction yields a locally unique real-analytic branch, with vanishing linear correction to the bifurcation parameter. Moreover, each finite tail critical value produces a simple local branch. These values accumulate at a parameter where the linearization ceases to be Fredholm. The accumulation point is nevertheless a bifurcation point in the usual topological sense. For $\sigma=1$, a
symmetry in the first
Fourier mode produces an exact two-mode vertical branch, including the exceptional non-Fredholm case $k=2$.
Read the full abstract ▾
Robotics
Probabilistic Scene Graphs: Hierarchical Representation and Real-time System
Waqas Ali, Michele Antonazzi, Timon Homberger et al. (+4 more)
3D scene graphs provide semantically rich and hierarchical representations for
robot perception. However, existing systems do not maintain uncertainty as an explicit belief or propagate it through the operations that construct and refine the graph. We introduce Probabilistic Scene Graph (PSG), a generalization of the conventional scene graph that represents a posterior over possible graphs, factorized into a discrete graph structure of entities, relations, and semantic attributes, and continuous states that ground them spatially, with uncertainty maintained over both components. Geometry is carried directly by the nodes rather than selected from a separately constructed metric map, so a metric map, where needed, follows from the graph rather than preceding it. We instantiate PSG's probabilistic spatial grounding with hierarchical graphs of Gaussians (HGG): each object primitive is represented by a full-covariance Gaussian under a Normal-Inverse-Wishart belief, and the same parametrization applied recursively within a node yields a geometry graph that resolves its surface at finer resolution. We then build a mapping pipeline that preserves these beliefs throughout graph construction and refinement: a purely graph-based coarse-to-fine
alignment registers observations by comparing node beliefs, while a nested Expectation-Maximization and factor-graph optimization jointly refines poses, object parameters, and internal geometry. Across six
datasets spanning indoor RGB-D, outdoor
LiDAR, and cross-modality deployment, HGG operates at sensor rate with near-constant
memory and achieves state-of-the-art object accuracy and zero-shot graph
alignment.
Read the full abstract ▾
All authors Waqas Ali, Michele Antonazzi, Timon Homberger, Thien-Minh Nguyen, Lukas Rosenberger Schmid, Patric Jensfelt, Yixi Cai
Logic
CIC + EM $\vdash$ Con(ZF): the consistency of ZF in type theory with excluded middle and no choice
Mario Carneiro
The sets-as-trees interpretation of set theory in a dependent type theory with an impredicative universe of propositions validates Zermelo set theory, and it validates Replacement if the type theory has a choice or description operator, which turns a functional relation into a function. It has been natural to expect that without such an operator the strength of the type theory drops well below that of $\mathrm{ZF}$. We show that it does not. In the type theory of Lean with two predicative universes, from excluded middle as the only assumption and with no axiom (no choice, no propositional extensionality, no quotients), we prove the consistency of $\mathrm{ZF}$, stated outright for a first-order proof system. The proof is formalized. The mechanism is the large elimination of the accessibility predicate over a type as large as the type of sets: a recursion on accessibility whose recursive calls are guarded by propositions, and whose later calls are indexed by the value of an earlier call, computes as a term any ordinal that is specified by a proposition through a well-founded tree of a certain shape. We give a rule that produces such a tree for every ordinal, unless some $V_\rho$ is already a model of $\mathrm{ZF}$; the rule does not choose a cofinal map into a limit ordinal but takes all definable ones at once. In the first case the sets-as-trees satisfy Replacement for arbitrary propositional relations. Either way $\mathrm{ZF}$ has a model. Finally, the double negation of excluded middle suffices, and what remains of it is exactly that membership is not not well-founded in the stable reading of sets; this in turn implies the double negation of Markov's principle.
Read the full abstract ▾
Artificial Intelligence
CraftBench-UE: Deterministic Evaluation for Coding Agents in Unreal Engine
Shutong Wu, Kevin Calderone and Andy Tsen
Building gameplay features in a game engine requires more than code, as code that compiles and runs does not necessarily implement the requested gameplay. We introduce CraftBenchUE, an evaluation harness that runs
agents in an isolated Unreal Engine environment, reconstructs their saved submissions in fresh projects, and applies deterministic build, asset, and runtime checks without an LLM judge. Based on the harness, we built a
benchmark consisting of 70 tasks spanning C++ source, Blueprint assets, and editor scripting. We evaluate seven models under two editor-tool configurations, with a file-and-shell baseline on C++ tasks. We further pair tasks that specify the same gameplay and use the same runtime tests, but require C++ and Blueprint as the deliverables. Across the 10 paired tasks, C++ completion rates exceed Blueprint by 30.0 and 42.9 percentage points in the two tool configurations. Among on-time Blueprint submissions in this paired set that pass asset checks, 42.2% and 50.0% fail explicit runtime assertions. These submissions satisfy asset requirements but fail the required gameplay tests. We will release the harness, task
benchmark, and our trajectory findings with the report.
Read the full abstract ▾
Plasma Physics
Real-Time Plasma State Prediction via FPGA-Accelerated Quantized Recurrent Probabilistic Neural Networks
Daniel Gaytan-Villarreal, Aiken Xie, Tu Pham et al. (+5 more)
Real time
plasma state estimation for control of Tokamak devices are challenging due to the stringent latency requirements of the
plasma control system (PCS). We present an end-to-end workflow for deploying a recurrent probabilistic neural network (RPNN) on FPGA hardware. We combine architecture size reduction with quantization-aware training via QKeras. The model is then synthesized using hls4ml, targeting a Xilinx Alveo U50 device. We report a design that fits comfortably within all four resource budgets (DSP, LUT, FF, BRAM) at deterministic sub-10~$\mu$s single-timestep latency, meeting the requirements for real-time inference inside a model-predictive-control-style
plasma control loop.
Read the full abstract ▾
All authors Daniel Gaytan-Villarreal, Aiken Xie, Tu Pham, Rohit Sonker, Chiara Amendola, Matteo Cremonesi, Cong Hao, Jeff Schneider
Astrophysics of Galaxies
Star formation and morphological trends in the Antlia cluster: Probing environmental influence out to 5R200
Ciria Lima-Dias, Antonela Monachesi, Gissel P. Montaguth et al. (+48 more)
Galaxy evolution in dense environments such as clusters is strongly affected by environmental processes that can alter both morphology and
star formation activity. We investigate these effects in the Antlia cluster out to 5R200 using Southern Photometric Local Universe Survey (S-PLUS) data. We derive H-alpha-based
star formation rates from the J0660 band and Sersic indices with GALFITM. Our analysis focuses on 154 spectroscopically confirmed
galaxies brighter than mr = 16. We classify
galaxies as early- or late-type systems using colour and Sersic index, and as quenched or
star-forming using specific
star formation rate. We examine their dependence on clustercentric distance, projected phase-space, substructure, and local density. We find that
star-forming
galaxies dominate from 1 to 5R200, while quenched
galaxies are more common within 1R200. Early- and late-type
galaxies show similar radial trends, with comparable fractions within 1R200 and a higher late-type fraction at larger radii. Quenched
galaxies dominate at log(Sigma10 [Mpc^-2]) >= 1.5, whereas late-type
galaxies dominate at log(Sigma10 [Mpc^-2]) <= 1.0. We identify substructures extending from the cluster centre to the outskirts. Substructures outside the central region are dominated by
star-forming and late-type
galaxies, with fractions similar to those of
galaxies not associated with detected substructures. However, when the massive central structure is included, the fraction of
star-forming
galaxies in substructures decreases, indicating a more environmentally processed population in the central region. Altogether, the presence of a massive central substructure together with additional substructures detected out to 5R200 indicates that Antlia is still undergoing mass assembly through group
accretion, consistent with a dynamically active and young cluster that is still assembling its
galaxy population.
Read the full abstract ▾
All authors Ciria Lima-Dias, Antonela Monachesi, Gissel P. Montaguth, Rodrigo F. Haack, Daniela E. Olave-Rojas, Hugo Méndez-Hernández, Sergio Torres-Flores, Luis Otárola, Juan Pablo Calderón, Analía V. Smith Castelli, Yara L. Jaffé, Christopher P. Haines, Diego Pallero, Vitor M. Sampaio, Cristóbal Sifón, Alvaro Alvarez-Candal, María Argudo-Fernández, Simón Véliz Astudillo, Maiara S. Carvalho, Letizia P. Cassarà, Cheng Cheng, Arianna Cortesi, Ricardo Demarco, Alexis Finoguenov, Marcos Fonseca-Faria, Thiago S. Gonçalves, Eduardo Ibar, Iván Lacerna, Elismar Lösch, Angela C. Krabbe, Ulrike Kuchner, Erik V. R. Lima, Amanda R. Lopes, Vitor H. Lopes-Silva, Amrutha B. Manjunatha, Natanael M. Cardoso, Sebastián Ortiz-Gómez, Gustavo B. Oliveira Schwarz, Umberto Rescigno, Boudewijn F. Roukema, Nicolás Tejos, Eduardo Telles, Benedetta Vulcani, Stephane V. Werner, Laerte Sodré Jr, Fábio R. Herpich, Felipe Almeida-Fernandes, Antonio Kanaan, Tiago Ribeiro, William Schoenell, Claudia Mendes de Oliveira
Computer Vision
QwenVLConnector: A Fast, Unified Medical VLM Chatbot for Fine-Grained Clinical Perception and Text Generation
Le Thien Phuc Nguyen, Thien Nguyen, Thanh-Huy Nguyen et al. (+3 more)
Most medical
vision-
language models (VLMs) excel at open-ended report generation and VQA but provide limited support for structured, fine-grained clinical perception within a unified interface. We present QwenVLConnector, a Qwen2.5-VL-based medical chatbot that unifies classification, multi-label classification, textualized detection, counting, regression, and free-form report generation under a single next-
token objective. Our key component is a lightweight dense multi-layer Connector that aggregates low- and high-level visual features, aligns them through the pretrained
vision Merger, and fuses them with the final visual representation without increasing sequence length. This design enriches visual
tokens with complementary spatial and semantic cues while preserving efficiency. On
FLARE-2D, QwenVLConnector improves detection F1 from 0.55 to 0.85, raises single-label classification from 0.37 to 0.51, and boosts report-generation GREEN by up to 18.3 points over the Qwen2.5-VL baseline. We further explore
multimodal in-context learning for report generation, showing additional improvements without updating model parameters. Overall, QwenVLConnector offers a unified and efficient framework for combining structured medical perception with open-ended clinical text generation. Our code can be found at https://github.com/plnguyen2908/QwenConnector.
Read the full abstract ▾
All authors Le Thien Phuc Nguyen, Thien Nguyen, Thanh-Huy Nguyen, Gia Minh Hoang, Anh Mai Vu, Ulas Bagci
History and Overview of Mathematics
Comparing different approaches to the concept of determinant: the multiplicative approach is the best!
Peteris Daugulis, Mikhail H Klin and Anita Sondore
The determinant is traditionally introduced through permutation expansions, multilinearity, or recursive expansion formulas - approaches that, while rigorous, often obscure its conceptual significance. Instead, an approach is advocated wherein the determinant is defined via its natural characterization as a multiplicatice map on
matrices, reflecting its role as a structural invariant of linear transformations under composition. This perspective defers combinatorial machinery until after the core properties are estaqblished, thereby lowering the cognitive barrier for students encountering the concept for the first time. The equivalence of this characterization with classical formulations is shown, and it is argued that this framework better aligns with the structural emphasis of m odern mathematical education.
Read the full abstract ▾
Systems and Control
TRACS: A Geometry-Aware Framework for Scalable Multi-Agent Path Finding in Warehouses
Siddhant Erande and Anuj Tiwari
Large scale warehouse automation relies on efficient multi
agent path finding (MAPF) to coordinate thousands of
robots in structured environments. Existing MAPF
algorithms primarily improve conflict resolution while representing warehouses as generic
navigation graphs, overlooking their inherent
geometric structure and traffic patterns. This paper presents TRACS (Traffic aware Routing and Aisle Coordination System), a geometry aware planning framework that exploits warehouse layout to simplify planning rather than introducing another conflict-resolution
algorithm. TRACS constructs a directed routing graph with alternating one way aisles that eliminates head on and edge swap conflicts by design, decoupling spatial routing from temporal traffic coordination. Independent hybrid graph grid routing is combined with lightweight edge based scheduling to avoid joint space time search while ensuring collision free execution. Experimental evaluation on warehouse
benchmarks against representative priority based, iterative repair, and search based MAPF planners shows that TRACS consistently achieves a 100% empirical success rate while substantially improving planning scalability. On fixed scene
benchmarks with up to 1000
robots, TRACS reduces planning time by up to 14.7X while maintaining competitive makespan, lower flowtime, and near optimal path quality. Under a fixed 10 minute planning budget, TRACS routes up to 5120
robots, roughly twice the largest fleet reached by the strongest baselines, while sustaining a 100% success rate, demonstrating the effectiveness of exploiting warehouse geometry for scalable robotic warehouse systems.
Read the full abstract ▾
Networking
Intent Interpretation at RIC Timescales: Jev Decision Models versus Large Language Models in 6G Open RAN
Delong Li, Xu Wang, Haochen Gong et al. (+2 more)
Intent-based Open RAN needs an interpreter that turns intents into A1 policies within the loop of the RAN intelligent controller (RIC). Decision models such as Jev-1.13.0 return typed policy fields, whereas generative large
language models (LLMs) produce the policy
token by
token. We ask whether the extra delay of LLMs costs control deadlines, RIC capacity, or
radio performance. We compare Jev-1.13.0 and two other decision models with LLMs on the RANIntent v1
benchmark, in closed-loop ns-3 simulation and on a real A1 and E2 path. Median interpretation takes 0.286 to 2.35 s, against under 25 ms for A1 and E2 transfer. Jev-1.13.0 meets the 1 s near-real-time budget on 99.8% of calls, while two hosted LLMs meet it on 17.9% and 0%. In the
radio network, ideal enforcement moves the affected-class service-level agreement (SLA) violation by 3.96 percentage points in the direction each intent requests, against no update at the base point. No hosted LLM showed a resolved increase over Jev-1.13.0 at that point. At the same point, per-second direct control gave no resolved SLA reduction over a numerical xApp. Slow interpreters miss the 1 s budget, and two interpreters saturate their queues at 2 intents/s, whereas no
radio penalty of slow interpreters was resolved at the base point.
Read the full abstract ▾
All authors Delong Li, Xu Wang, Haochen Gong, Rui Lang, Guangsheng Yu
Human-Computer Interaction
Evaluative Dynamics of AI Integration and Expert Performance under Epistemic Dependence across Heterogeneous Stakes
Dennis Kim, Roya Daneshi, Nikhil Krishnaswamy et al. (+2 more)
AI is increasingly integrated into expert workflows, yet how integration affects perceptions of the expert, AI, and their combination remains unclear in domains where lay users are epistemically dependent on AI-assisted experts. We examine this through a novel controlled medical study (N = 166) and a direct cross-domain analysis with pre-existing academic-advising data (n = 157, combined N = 323). Expert errors reduced evaluations of the human expert across domains. Perceived expertise, however, varied by AI integration strategy in the higher-stakes medical task, where automatic AI oversight produced higher ratings than expert-only or expert-initiated AI. Exploratory ordinal sensitivity analyses identified a performance-contingent reuse pattern, with automatic oversight producing greater intended reuse after successful medical performance. Overall, performance-related recalibration appeared comparatively portable, while integration-structure effects were more selective and context-sensitive. These findings suggest that system designers should consider how AI enters expert workflows, not only whether it is present.
Read the full abstract ▾
All authors Dennis Kim, Roya Daneshi, Nikhil Krishnaswamy, Bruce Draper, Sarath Sreedharan
Chemical Physics
ECENet: An Edge Cluster Expansion Line-Graph Neural Network
R. Allen LaCour and Teresa Head-Gordon
Machine-learned interatomic potentials (MLIPs) have emerged as a promising alternative to classical force fields and first-principles theory for predicting the properties of chemical and material systems. Many MLIPs are graph neural networks with O(3)-equivariant features, whose accuracy comes at substantial computational cost. Here we introduce the edge cluster expansion (ECE), an analogue of the atomic cluster expansion in which the environment is expanded around edges between atom pairs rather than single atoms, and build upon it to develop the line-graph neural network ECENet. ECENet uses O(2)-equivariant features that persist on the edges between atoms, giving it natural access to O(2) operations that are cheaper and less restrictive than their O(3) counterparts. ECENet performs at the state of the art on the MD22
benchmark and lies on the accuracy-cost Pareto frontier when trained on the SPICE-MACE-OFF
dataset. Furthermore, with long-range electrostatics implemented via latent Ewald summation, ECENet accurately predicts molecular dipole moments and the infrared spectrum of
liquid water. The frontier ECENet architecture establishes equivariant edge-centered representations as an efficient and physically expressive foundation for equivariant MLIPs.
Read the full abstract ▾
Robotics
AquaCap: A Training-Free Underwater Embodied Agent with Code-as-Policy
Xiaoshi Li, Yule Xu, Chunghiu Kong et al. (+5 more)
Recent advances in
vision-
language-action models have stimulated growing interest in
underwater embodied intelligence. However, their reliance on large-scale interaction data limits their applicability
underwater, where data collection is costly and scarce. To address this challenge, we present AquaCap, a training-free Code-as-Policy framework for autonomous
underwater navigation and manipulation. AquaCap employs a dual-layer
agent that translates task instructions and environmental observations into condition-aware plans and executable control programs. Structured perception then provides the
agent with semantic,
geometric, and reliability-aware observations under degraded
underwater conditions. A failure-aware
memory diagnoses unsuccessful actions and supports closed-loop replanning and code revision. This design enables online adaptation without task-specific training or parameter updates. AquaCap achieves a 66.43% success rate in simulation. Real-world experiments further demonstrate autonomous grasping and object transport with an ROV, including the manipulation of targets displaced by hydrodynamic disturbances.
Read the full abstract ▾
All authors Xiaoshi Li, Yule Xu, Chunghiu Kong, Yizhou Zhou, Yang Liu, Hao Yang, Zihao Huang, Yunxiao Shan
Robotics
Tip Manipulation in Soft Everting Robots via Wall Retraction and Deployable Fingers
Nelson Badillo Perez, Niccolo Pagliarani, Matteo Cianchetti et al. (+1 more)
Soft everting
robots can traverse long, confined paths by continuously growing, yet active interaction remains limited to a single tool fixed at or near the tip, unable to be repositioned on demand and difficult to reconcile with the
robot's soft body. We introduce a tip-manipulation and multi-tool deployment strategy for soft everting
robots based on wall retraction, implemented with a base roller assembly that independently meters membrane flow in the outer wall while a tail spool regulates growth in the internal tail section. Coordinated wall and tail actuation decouples
robot length from membrane-material position, enabling membrane-mounted devices to be transported, exposed, and repositioned at selected locations near the distal tip. We pair this capability with ultralight pleated inflatable fingers integrated into the membrane, fabricated from TPU-coated nylon with an internal airtight bladder. The fingers achieve large bending at low pressures (approximately 100 degrees at 50 kPa in high-pleat designs) and generate blocking forces up to 1.9 N, while remaining limp during transport. The system demonstrates adaptive grasping across diverse household objects (21 to 550 g; 16 to 200 mm), three-dimensional object manipulation and stacking, environmentally braced extension, distal camera panning for confined-space inspection, and controlled sequential payload delivery. These results enable
embodied and reversible tip manipulation for soft growing
robots in cluttered and tortuous environments.
Read the full abstract ▾
All authors Nelson Badillo Perez, Niccolo Pagliarani, Matteo Cianchetti, Robert D. Howe
Robotics
Selective Commitment for Language-Guided Object Retrieval under Partial Observability
Wonhee Koh, Sushil Samuel Dinesh, Hansol Ko et al. (+2 more)
Language-guided object
retrieval under partial observability requires deciding whether to gather more evidence, interact with the scene, grasp a candidate, or abstain. We present a closed-loop framework that coordinates these decisions for retrieving a target specified in relation to a reference container. The framework maintains a persistent joint belief over target identity, container relation, and presence through tracked-object, unobserved-target, and target-absent hypotheses. View-conditioned categorical VLM observations update this belief; conformal grasp eligibility and
robot feasibility govern commitment, while finite-horizon belief-space planning selects information-gathering actions. Across five different scenarios, our proposed method succeeds in 19/25 simulation episodes versus 12/25 for the best-performing task-adapted baseline and is the only evaluated policy to achieve at least one success in each scenario. Ablations show that cross-view
memory improves success under partial occlusion, while the full system does not consistently outperform simplified variants. Real-
robot trials demonstrate closed-loop re-observation and autonomous recovery from injected grasp failures, while injected viewpoint failures end in false defer. Experimental results demonstrate the feasibility of coordinating evidence gathering and selective grasp commitment within a unified framework for
retrieval under partial observability.
Read the full abstract ▾
All authors Wonhee Koh, Sushil Samuel Dinesh, Hansol Ko, Shinkyu Park, Eungjoo Lee
Artificial Intelligence
From Inference Engine to Inference Control Plane: Connecting vLLM, llm-d, and the Evolution of Efficient Distributed LLM Serving
Twinkll Sisodia
Large
language model (LLM) inference is evolving from an engine-local optimization problem into a distributed control problem involving reusable state, phase placement, heterogeneous accelerators, networking, autoscaling, reliability, and service-level objectives. This paper connects that transition across peer-reviewed systems research, open-source implementations, and documented production studies. It treats vLLM and llm-d as complementary layers: model-serving engines optimize execution through mechanisms such as PagedAttention, continuous batching, kernels, quantization, and parallelism, while an inference control plane can optimize where, when, and under what policy execution occurs across a fleet. The contribution is synthesis rather than a new
benchmark; all reported performance and deployment results remain attributed to their original sources. The combined evidence suggests that the scarce resource in modern inference is shifting from raw FLOPs alone toward managed state, placement, network movement, reliability, and decision quality. We propose an Inference Execution Planner that selects feasible execution plans rather than only endpoints, including aggregated versus disaggregated
topology, KV source and transfer action, hardware variant, routing/admission policy, and slower scaling decisions. We also provide a source-local
benchmark atlas, a bottleneck-migration taxonomy, practical deployment guidance, an evaluation framework based on SLO-goodput, and research questions for agentic,
multimodal, heterogeneous, and resilient inference.
Read the full abstract ▾
Mesoscale and Nanoscale Physics
Magnetic Quadrupole Lens and Current-Induced Quadrupolar Dynamics of Antiskyrmion
Linfeng Du, Yang Liu, Peng Yan et al. (+1 more)
We demonstrate that antiskyrmions can serve as
magnetic quadrupole lenses that focus or defocus electric currents traversing them. We showcase the antiskyrmion core as a focal point where electron beams converge for specific injection directions, leading to anisotropic and nonuniform current distributions. Thus the current-induced antiskyrmion dynamics is profoundly modified by the
magnetic quadrupole
lensing via the
spin-transfer torque. By incorporating
quantum transport and micromagnetic simulations iteratively, we self-consistently capture the coupled dynamics of conduction electrons and antiskyrmions. Our results unveil a quadrupolar antiskyrmion
Hall effect that is highly anisotropic with respect to the current direction and can be effectively tuned by the
spin-orbit coupling. In particular, the current focusing (defocusing) dramatically enhances (suppresses) the antiskyrmion speed while simultaneously reducing its
Hall angle, thereby overcoming a key challenge for antiskyrmion racetrack
memories. Our work establishes the intrinsic current
lensing as a powerful mechanism for manipulating topological
magnetic textures and opens new avenues for antiskyrmion-based spintronics.
Read the full abstract ▾
All authors Linfeng Du, Yang Liu, Peng Yan, Ying Su
Combinatorics
There Are No Nontrivial Perfect Codes in the Johnson Scheme
Tuvi Etzion, Xiande Zhang and Wenjie Zhong
In his seminal work from 1973, Delsarte conjectured that the Johnson schemes admit no nontrivial perfect codes. We prove this fundamental
conjecture which is central in combinatorial coding theory. Our arguments combine lower bounds on the design strength of a perfect code with divisibility conditions for the size of a metric ball. We show that the ball size divides the greatest common divisor of specified binomial coefficients. An exact formula for this gcd implies squarefreeness and localizes a power of each
prime divisor of the ball size to a short interval. Lloyd
polynomials, explicit
prime estimates, and bounds for linear forms in logarithms then exclude all sufficiently large radii. Additional arithmetic arguments handle the remaining radii.
Read the full abstract ▾
Machine Learning
Provably Efficient Reinforcement Learning in Continuous-Time Episodic MDPs with Poisson Decision Epochs
Kenny Guo, Valentio Iverson, Sahan Wijetunga et al. (+1 more)
Many real-world reinforcement learning (RL) problems evolve in continuous time, where decisions occur at irregular, event-driven intervals rather than at fixed discrete steps. We study episodic continuous-time Markov Decision Processes (MDPs) in which decision epochs are governed by a homogeneous Poisson process and the reward and transition dynamics vary smoothly over time. We consider both a fixed number of jumps per episode and a fixed time budget with a random number of Poisson decision epochs. Under a Lipschitz continuity assumption in time, we exploit local smoothness through discretization and extend both UCRL (Auer and Ortner 2006) and Q-learning (Jin et al. 2018) to this setting, proving $\widetilde{O}(T^{2/3})$ regret bounds for both model-based and model-free
algorithms. Finally, we establish matching $\widetilde{\Omega}(T^{2/3})$ minimax lower bounds, showing that the rate is optimal up to logarithmic factors. These results provide the first tight regret guarantees for Lipschitz-smooth continuous-time episodic MDPs with Poisson decision epochs.
Read the full abstract ▾
All authors Kenny Guo, Valentio Iverson, Sahan Wijetunga, William Chang
Quantum Physics
Performance optimization of cascaded traveling wave Josephson parametric amplifiers
Ilari Lilja, Ekaterina Mukhanova, Stanislav Khaldeev et al. (+3 more)
Traveling-wave parametric amplifiers (TWPAs) based on Josephson metamaterials provide broadband gain with near-
quantum-limited added noise. Whereas long nonlinear metamaterial devices can deliver high gain, short arrays suffer less from dissipation, pump depletion, and internal standing waves which can degrade noise performance. Here, we demonstrate a cascaded TWPA architecture that combines the advantages of both approaches by employing a short (736-element), low-dissipation Superconducting Nonlinear Asymmetric Inductive eLement (SNAIL)-based TWPA as the first amplification stage, followed by a conventional long (1632-element) TWPA that provides additional gain. The resulting amplifier cascade achieves nearly 30 dB of total gain over a tunable bandwidth of approximately 1 GHz while maintaining added noise close to the
quantum limit. Our results establish a cascade of TWPAs as a practical approach for high-gain, broadband,
quantum amplification with applications in
quantum information processing, multi-mode
entanglement, and
quantum sensing applications at microwave frequencies.
Read the full abstract ▾
All authors Ilari Lilja, Ekaterina Mukhanova, Stanislav Khaldeev, Ilya Golokolenov, Visa Vesterinen, Pertti Hakonen
Machine Learning
Perplexity Cost Understates What Activation Quantisation Breaks
Anish Sathyanarayanan
Activation quantisation is usually evaluated with an aggregate metric, perplexity, averaged over every
token a model predicts. We ask whether that average identifies which computations a quantiser damages. Perplexity turns out to be a reliable aggregate signal: across 12 models from four families and 780 within-model comparisons, the arm perplexity prefers also retains more induction and more
retrieval in all but 2.1 and 4.0 percent of cases respectively. But where perplexity has risen by only a factor of 1.2 to 1.5, induction still keeps 0.959 of its intact accuracy while
retrieval has already fallen to 0.554, a gap the aggregate number does not surface. This gap has structure, not just size: a matched Gaussian-noise control of the same per-channel magnitude leaves it largely intact, and randomising only the sign of the quantisation error, every magnitude held fixed, is nearly as harmless, so magnitude alone does not explain the damage. Quantising in a rotated basis, which changes coordinate
alignment without changing error magnitude, restores induction from 0.001 to 0.980 at three average bits per
token in a single-block intervention, though
retrieval recovers less completely at the same setting (0.694); end-to-end at four average bits, induction reaches 0.968 and
retrieval 0.534. The pattern holds on two further models up to 32B parameters and, in the deployed configurations we tested, under AWQ once activations are pushed to 4 bits. A perplexity target bounds the average cost of a transformation applied to the activation; it does not, by itself, show which computations survived.
Read the full abstract ▾
Algebraic Geometry
A polynomial formula for the Albanese map of the Cartwright-Steger surface
Zoltán Szabó
The Albanese fibration of the Cartwright–Steger surface is a genus-$19$ Lefschetz fibration over a
torus. We give an explicit formula for it: in the projective model of Borisov and Yeung, the map is represented by three homogeneous
polynomials of degree $13$ with integer coefficients, and its target is the
elliptic curve $Y^2Z=X^3-3888Z^3$. A recognition
theorem shows that it suffices to verify one
polynomial identity on the surface. A coefficient bound allows us to establish the identity over $Q$ from computations modulo
primes. We also determine exact coordinates for the three nodes and their images on the
elliptic curve.
Read the full abstract ▾
Systems and Control
Composite Adaptive Higher-Order Control Barrier Functions for Joint Frequency-RoCoF Safety in Low-Inertia SIDS Microgrids
Shankar Ramharack, Jonathan Nancoo and Rajiv Sahadeo
Small Island Developing States (SIDS) face simultaneous frequency nadir and rate-of-change-of-frequency (RoCoF) violations under high renewable penetration. This paper proposes a Composite Adaptive Higher-Order Control Barrier Function (CA-HOCBF) for BESS-coupled virtual synchronous generators that jointly enforces IEEE~1547 frequency ($\pm0.8$\,Hz) and ENTSO-E RoCoF ($\pm1$\,Hz/s) limits as hard safety constraints under parametric uncertainty. The method unifies three tools: (i)~a dual barrier architecture with a zeroing CBF for frequency and an algebraic RoCoF constraint with a robust inertia-floor fallback; (ii)~a composite energy function coupling logarithmic safety barriers with quadratic parameter and disturbance error terms, which drives safety-weighted adaptation; and (iii)~a disturbance observer integrated into the barrier dynamics. We prove that the closed-form QP is always feasible under a BESS capacity condition and, under explicitly stated assumptions, that the composite energy remains finite along trajectories, rendering the joint safe set forward invariant without requiring parameter convergence. Simulation on a 10\,MW Caribbean microgrid ($H=2$\,s, 70% renewables) shows the CA-HOCBF eliminates all frequency and RoCoF violations under a compound disturbance, achieving steady-state error of 0.001\,Hz and RoCoF of 0.50\,Hz/s, where fixed-gain VSGs suffer a 0.62\,Hz offset with 2.25\,Hz/s RoCoF violations.
Read the full abstract ▾
Earth and Planetary Astrophysics
Searching for Embedded Protoplanets with the Keck/NIRC2 Vortex Coronagraph: Confirmation of a Core-accretion Planet in the Narrow Gap of the Elias 2-24 Disk
Andrea Bernardi, Alice Zurlo, Lucas A. Cieza et al. (+8 more)
We present high-contrast imaging observations of seven
stars obtained with the Keck/NIRC2
vortex coronagraph between 2018 May and June in the $L'$ and $M_s$ bands. All targets host protoplanetary disks with substructures such as gaps and cavities, potentially tracing ongoing
planet formation. Our aim is to assess whether these embedded
planets can be directly detected. We reduced the observations with a custom pipeline based on the VIP Python package. Postprocessed images were obtained using angular differential imaging and reference differential imaging combined with principal component analysis. Finally, 5$\sigma$ contrast curves and detection limits were derived for all targets. We confirm the planetary candidate around Elias 2-24, previously proposed in the literature, at a separation of 394$\pm$31 mas (54.9$\pm4$.3 au) and a position angle of 298.8$\pm$3.2$^\circ$, within the narrow gap of the protoplanetary disk. Comparing the derived photometry with 1 Myr isochrones, we infer a mass between 1.9 and 4.0 Jupiter masses for Elias 2-24 b, without considering
accretion effects, in agreement with estimates considering the properties of the gap and
planet-disk interactions. For all the systems in the survey, the 5$\sigma$ detection limits reach sensitivities down to 1-11 Jupiter masses at separations $\gtrsim$0.25". Since Elias 2-24 b (age $\lesssim$1 Myr) is the youngest
exoplanet discovered to date, its confirmation has important implications for
planet formation timescales and the origin of the gaps observed in very young disks.
Read the full abstract ▾
All authors Andrea Bernardi, Alice Zurlo, Lucas A. Cieza, Garreth Ruane, Valentin Christiaens, Anuroop Dasgupta, Greta Guidi, Dimitri Mawet, Dino Mesa, Sebastián Pérez, Jonathan P. Williams
Computer Vision
MM-ContextFold: Context Folding for Multimodal Agentic Retrieval
Yang Tian, Fan Liu, Jingyuan Zhang et al. (+3 more)
Multimodal Agentic
Retrieval (MAR) requires
agents to solve complex information-seeking tasks by iteratively invoking external tools. Typical frameworks such as ReAct maintain raw
multimodal inputs and the accumulating interaction history in a single, ever-growing context, leading to the context explosion problem. While existing methods alleviate this issue by compressing redundant text, effective strategies for managing
token-intensive visual content remain largely underexplored. To address this gap, we first conduct a systematic empirical study of approximately 10,000 trajectories. The results show that as visual cues are progressively extracted through external tools and textualized into the context, raw images become increasingly redundant. Continued image retention is associated with higher output
entropy and can even degrade task accuracy. Motivated by these findings, we propose MM-ContextFold, a training-free framework that loads raw images only when needed. It maintains a persistent, text-only main context for high-level planning and spawns ephemeral branch contexts for image-dependent subtasks. Within each branch, the
agent loads the relevant images, completes the subtask, and folds the result back into the main context as a concise textual summary; the images and branch trace are then discarded. Experiments on seven MAR
benchmarks across five backbone models show that MM-ContextFold improves average accuracy by 6.3 percentage points over ReAct while reducing the working context length by 27.5%.
Read the full abstract ▾
All authors Yang Tian, Fan Liu, Jingyuan Zhang, Zhenyang Li, Yupeng Hu, Liqiang Nie
Statistical Methodology
Modeling bounded well-being indices using Bayesian double generalized beta regression with spatial and temporal borrowing
Abhi Jain, Michael LaValley, Kimberly A. Dukes et al. (+1 more)
Health and well-being indices are widely used to assess population health outcomes and inform policy decisions. Individual-level assessment of well-being can be used to develop community-level indices that measure wellness for different geographical units. While many existing indices operate at coarse geographic levels such as counties or states, finer spatial resolution can offer more actionable insights. We present a novel
Bayesian double generalized beta regression framework to model a bounded individual-level well-being index (WBI) using annual survey data collected from 2021 to 2023 in Massachusetts. Although survey respondents may differ across years, responses are geotagged to ZIP Code Tabulation Areas (ZCTAs), enabling the integration of both spatial and temporal information. Our framework incorporates spatial dependencies via a graph
Laplacian matrix that encodes driving time-based ZCTA neighborhood structure, and leverages temporal borrowing by using posterior spatial effect estimates from one year to inform priors in the next. This dual-borrowing strategy within a
Bayesian double generalized beta regression framework enhances estimation precision, particularly in areas with sparse data, and improves inference for smaller geographic units. We demonstrate the utility of our method through a realistic simulation study that highlights improved estimation when borrowing spatial and temporal information. In the real data analysis, we model individual-level well-being for residents in Massachusetts and find that income, education status, and marital status are most associated with WBI. Additionally, we observe that ZCTAs in Western Massachusetts, Cape Cod, and those near Boston perform best with the highest spatial effects.
Read the full abstract ▾
All authors Abhi Jain, Michael LaValley, Kimberly A. Dukes, Shariq Mohammed
Mathematical Physics
Long time stability of low–frequency packets of modes at finite total energy
Dario Bambusi and Antonio Ponno
We consider the FPUT $\alpha$-model with $N$ particles and initial data with small total energy $E$, independent of $N$. In particular, we prove that if the energy is initially given to
Fourier modes of wave number $\pi k/N\leq \theta_*$, then it does not flow to modes of wave number larger than $\theta_*$ up to long times of order $E^{-(1-\epsilon/2)}$, $\theta_*$ and $\epsilon$ being small numbers. The proof relies on the closeness of the Toda
lattice to the FPUT system. Unlike previous approaches, we do not make use of the action-angle variables of the Toda system, but we exploit the Flaschka integrals of the latter. Actually, we find a suitable recombination of the Flaschka integrals whose quadratic part is used to control the packet, the higher order part fulfilling very good estimates.
Read the full abstract ▾
Robotics
Verti-WM: A Physics-Aided Exteroceptive World Model for Off-Road Reinforcement Learning
Chenhui Pan, Tong Xu and Xuesu Xiao
Reinforcement learning for off-road
navigation requires extensive vehicle-terrain interaction data, which are costly to collect in high-fidelity simulation. World models offer a promising alternative by replacing simulator roll-outs during policy optimization. However, an off-road world model must condition state transitions on exteroceptive terrain information, which proprioception alone does not provide. This challenge is further amplified by the need to model both rigid and deformable terrain, where data-driven and physics-based approaches offer complementary strengths. We propose Verti-WM, a physics-aided exteroceptive world model that recurrently fuses a frozen
Transformer for rigid terrain and a neuro-symbolic terramechanics model for deformable terrain. Elevation and semantic observations queried from a supplied map at each predicted pose condition fusion, enabling six-degree-of-freedom rollouts for policy optimization without further simulator access. Verti-WM reduces prediction error by 34.6% and 21.7% over data-driven and physics-based baselines, respectively. Policies trained entirely within Verti-WM achieve comparable task success rates while reducing computation time by 23.6X relative to direct training in the high-fidelity simulator. We further validate Verti-WM using real-world data, enabling policy optimization within learned real-world kinodynamics and achieving a 80% success rate on the Verti-4-Wheeler platform, compared with 40% for direct sim-to-real transfer.
Read the full abstract ▾
Machine Learning
Whitening Inverts the Hierarchy: What the Norm of a Whitened Embedding Measures
Mohammed Ahnouch and Lotfi Elaachak
Whitening a foundation-model
embedding and using its squared norm as a training-free likelihood surrogate is motivated by the observation that whitened coordinates often appear approximately standard normal. We show that this observation follows from the projection central limit
theorem and therefore does not imply a Gaussian joint distribution. Across multiple encoders and three training objectives, we find systematic over-dispersion of the whitened radius relative to the Gaussian reference, including against distributional clones with identical mean and covariance. We further show that the commonly reported agreement between empirical and theoretical norm statistics is an algebraic consequence of in-sample whitening and does not constitute evidence for Gaussianity. We identify the mechanism behind this behavior: whitening reverses the encoder's spectral hierarchy, shifting the contribution to the squared norm toward near-degenerate directions that encode predominantly noise. In these directions, the dominant variability is governed by a single input-dependent scale. We estimate this scale from two moments and use it to predict, without additional free parameters, the cross-dependence between disjoint spectral halves. These results indicate that the squared whitened norm is better interpreted as a Mahalanobis measure of semantic atypicality than as a log-likelihood. This interpretation explains both its practical effectiveness and its calibration failures: the statistic can rank and detect atypical samples consistently with nonparametric density estimates and across encoders trained with different objectives, while Gaussian tail thresholds can be inaccurate by orders of magnitude. etc.
Read the full abstract ▾
Hardware Architecture
Benchmarking StreamNTT with a Verilog-to-Routing Toolchain
Wei He, Young-kyu Choi, Hyunwoo Park et al. (+1 more)
As post-
quantum cryptography
algorithms move toward large-scale data center deployment, hardware acceleration of their computational bottleneck, which is the number theoretic transform (NTT), has gained increasing
attention. StreamNTT, a high-level synthesis- and field-programmable gate array-based accelerator, achieves state-of-the-art throughput through various optimization techniques. However, its reliance on a commercial tool and a device makes direct comparisons difficult for researchers without access. We address this by building StreamNTT on an open-source Verilog-to-Routing toolchain, which achieves similar digital signal processing and multiplier usage. Significant differences in internal
memory utilization indicate that further
memory-level optimization is needed to approach commercial tool performance.
Read the full abstract ▾
All authors Wei He, Young-kyu Choi, Hyunwoo Park, Sunwoong Kim