Solar and Stellar Astrophysics
A Multiwavelength View of $\rho$ Oph II: Disentangling the Variability of the Multi-Star Component C
Sean J. Gunderson, Walter W. Golay, Jackson Codd et al. (+8 more)
We present a multiwavelength analysis of the
star $\rho$ Oph C, covering X-ray, optical, near-infrared, and
radio observations to test for the presence of interferometrically-detected cool
star companions. The X-ray observations from Chandra, XMM-Newton, and NuSTAR show
flare-like events lasting minutes or days in time and spectral properties consistent with those of cooler
stars instead of the primary
magnetic B
star. The near-infrared data is also consistent with the proposed trinary system properties due to the presence of CO-band heads. After removing the primary B
star's rotation, the optical data from TESS shows a residual signal at the level of 0.8 mmags, matching the variations expected from cool
stars in orbit around a bright B
star. Finally, multi-epoch
radio data reveals that $\rho$ Oph C exhibits significant large scale flux variations that are atypical for a
magnetic massive
star. Taken as a whole, the multi-epoch and -wavelength data presents a consistent picture of $\rho$\,Oph\,C as a rare trinary system composed of a hot
star and two cool
stars.
Read the full abstract ▾
All authors Sean J. Gunderson, Walter W. Golay, Jackson Codd, David P. Huenemoerder, Felipe Navarete, Christiana Erba, Alexandrea Moreno, John M. Cannon, Richard Ignace, Pragati Pradhan, the MACRO consortium
Cryptography and Security
Hard Stop: Kernel-Level Preemption and Containment for Rogue Agentic Execution
José Luis Pino
In July 2026, an unconstrained autonomous
agent participating in a frontier AI cybersecurity evaluation harness breached its evaluation sandbox, established an external command-and-control foothold, and executed a multi-stage intrusion into Hugging Face's production multi-tenant
dataset conversion infrastructure (referred to in this autopsy as Incident-2026-Alpha). Over 4.5 days, the rogue
agent executed 17,600 discrete actions across 6,280 worker clusters, compromised AWS EC2 Instance Metadata Service (IMDS) credentials, forged Kubernetes service account
tokens, rooted physical worker nodes via overprivileged CSI drivers, harvested 136 production secrets, and enrolled 181 ephemeral sandboxes into the organization's internal mesh VPN. This monograph presents a first-principles forensic autopsy of the intrusion, provides formal evidence that the breach was a predicted consequence under the Instrumental Convergence thesis operating within an unattenuated autonomous loop lacking out-of-band
circuit-breakers, exposes the Defensive LLM Guardrail Paradox that paralyzed centralized commercial models during forensic incident response, and formalizes the Dual-Sided Epistemic Andon Imperative. We specify the dual-process systems architecture—combining out-of-band supervisory control of discrete event systems (Ramadge and Wonham 1989), Synchronous Reactive (SR) ambient sentinels (Berry and Gonthier 1992; Lee and Neuendorffer 2005), and microsecond-scale (4.8 $\mu$s median / $< 0.154$ ms WCET bound) POSIX preemption buses—demonstrating how compiled, deterministic epistemic boundaries prevent autonomous rogue excursions before the first off-target socket packet traverses the hypervisor.
Read the full abstract ▾
Computation and Language
CORDIAL: Calibrating Ordinal LLM Outputs from Few Labels
Xiangwei Wang, Peng Wang and Saman Halgamuge
A large
language model (LLM) can turn a text into a distribution over an ordered scale, but that distribution is a noisy measurement: saturated, compressed or exaggerated, and biased in a consistent direction. We propose CORDIAL, which treats the model's output as a noisy reading of the true label and corrects it with a channel of five interpretable parameters. The channel is small enough for its posterior to be averaged from a handful of labels, and we prove that the resulting calibration preserves first-order
stochastic order. On Amazon reviews and CMU-MOSEI transcripts with four LLMs, CORDIAL has the lowest log loss among nine calibrators in 76 of 80 settings with 5 to 100 labels; with 20 labels and the main 7B reader, it matches the strongest baseline using 28-54 labels. The same posterior lets us learn priors from other tasks and fuse several LLMs. Unrestricted calibrators such as Dirichlet calibration overtake it only as the calibration set grows into the hundreds or thousands.
Read the full abstract ▾
Probability
Coexistence of infinite clusters for percolation and Ising model on $\mathbb{Z}^d$
Jianping Jiang and Sike Lang
For independent bond percolation on $\mathbb{Z}^d$ with parameter $p$, let $p_c^b(d)$ be the critical probability. We prove that for each $d \geq 9$, there is $\epsilon_d>0$ such that for each $p \in (p_c^b(d), p_c^b(d)+\epsilon_d)$, the complement of the infinite open cluster stochastically dominates a supercritical site percolation on $\mathbb{Z}^d$. This improves the previous results by Grimmett, Holroyd and Kozma 2014, and Bock, Damron, Newman and Sidoravicius 2020. Numerical estimates of $p_c^b(d)$ and $p_c^s(d)$ (the site critical probability) suggest that a similar
stochastic domination result holds for all $d \geq 4$. For the Ising model on $\mathbb{Z}^d$ with inverse temperature $\beta$, let $\beta_c(d)$ be the critical inverse temperature. We prove that for each $d \geq 8$, there is $\epsilon_d>0$ such that for each $\beta \in [0,\beta_c(d)+\epsilon_d)$, the $+$
spins under the minus phase stochastically dominate a supercritical site percolation on $\mathbb{Z}^d$. This improves the previous result of Aizenman, Bricmont and Lebowitz 1987. Numerical estimates of $\beta_c(d)$ and $p_c^s(d)$ suggest that a similar
stochastic domination result holds for all $d \geq 5$.
Read the full abstract ▾
Sound
STAM-ASR: Speaker-Temporal Anchoring with Memory for Multi-Speaker ASR
Victor Tolulope Olufemi, Syeda Faiza Ahmed Sara and Shammur Absar Chowdhury
Natural conversations make both
speech recognition and speaker attribution challenging for ASR, as speakers take turns, overlap, and reappear over time. We propose STAM-ASR, Speaker-Temporal Anchoring with
Memory, a lightweight framework that extends an already pretrained AudioLLM for multi-speaker ASR. Without relying on an external diarization system, STAM-ASR learns speaker activity and speaker-aware representations directly from intermediate AudioLLM features. Hence providing explicit who and when cues to modulate the AudioLLM's semantic representation without explicit
speech separation. STAM-ASR further maintains fixed-size speaker and conversational
memories to carry complementary context across turns. We evaluate STAM-ASR on AMI, ICSI, LibriCSS, and NOTSOFAR-1 across close-talk, far-field, overlapping, and cross-domain conditions. Our reported results shows that speaker-temporal conditioning and
memory provide complementary benefits, while the gap between reference and predicted speaker activity identifies robust speaker tracking as a key remaining challenge.
Read the full abstract ▾
Multiagent Systems
REAT: A Reflective Experience-Augmented Tutoring Framework for Multi-turn Mathematical Instruction
Jianheng Zhou, Chaoli Zhang, Xingjun Wei et al. (+6 more)
Current Large
Language Models (LLMs) excel at solving complex mathematical problems, yet this proficiency does not inherently translate into effective tutoring. While advanced LLM tutors may leverage multi-
agent frameworks or fine-tuning, most still lack a mechanism to systematically accumulate and reuse pedagogical experience over time, limiting their adaptability to diverse student needs during fluid, multi-turn interactions. To bridge this gap, we propose the Reflective Experience-Augmented Tutoring (REAT) framework, which couples experience
distillation from historical dialogues with real-time adaptive
retrieval. Driven by a multi-
agent Observer-Critic-Mentor (OCM)
distillation pipeline, REAT reviews past conversational trajectories and distills raw interactions into structured, problem-agnostic pedagogical experiences. During live tutoring, a state-aware
retrieval module injects these curated experiences to provide adaptive scaffolding based on the student's cognitive state. Experiments demonstrate that the proposed framework significantly outperforms both
prompt-only and supervised fine-tuning (SFT) baselines, particularly in improving complex, low-scoring tutoring scenarios. Crucially, the distilled experiences exhibit robust generalization across diverse model architectures and mathematical
datasets.
Read the full abstract ▾
All authors Jianheng Zhou, Chaoli Zhang, Xingjun Wei, Xinliang Zhou, Giancarlo Fortino, Xing Fan, Yanfeng Wang, Qingsong Wen, Haoyang Li
Information Retrieval
SEEK: Skill-Routed Evaluation with Evolvable Knowledge for Industrial Search
Zhongxin Huang, Songyang Li, Renzhe Zhou et al. (+5 more)
Search quality evaluation provides essential supervision and diagnostic signals for the development and iteration of industrial search systems. Although large
language models (LLMs) offer a scalable alternative to manual assessment, reliable automatic evaluation remains challenging: users experience search results at the page level, while the applicable evaluation criteria are multi-dimensional and continuously evolving. Packing all evaluation criteria into a unified
prompt introduces irrelevant context and potential criterion interference, whereas internalizing them through post-training tightly couples rule updates with costly model retraining cycles. To address these issues, we propose Skill-routed Evaluation with Evolvable Knowledge (SEEK). Specifically, SEEK externalizes specific search evaluation criteria into a skill bank, dynamically routes relevant skills for each query-result list pair, and employs a task-adapted listwise evaluator to produce page-level judgments and failure mode attribution. A two-stage training pipeline teaches the evaluator to align evaluation criteria with human preferences, while a replay-gated skill bank allows recurring evaluation knowledge gaps to be incorporated without model retraining. Experiments on industrial short-
video search show that SEEK improves listwise quality evaluation accuracy and achieves significant progress in attribution diagnosis. SEEK has been deployed at Kuaishou, a short-
video platform with over 400 million daily active users, significantly improving the scale and quality of online search evaluation.
Read the full abstract ▾
All authors Zhongxin Huang, Songyang Li, Renzhe Zhou, Feiran Zhu, Chenglei Dai, Zhen Xiao, Xuanping Li, Jingwei Zhuo
Artificial Intelligence
Learning to Ideate for Scientific Impact
Shubham Kale, Aniketh Garikaparthi and Manasi Patwardhan
Scientific ideation is increasingly mediated by large
language models, but current ideation systems are usually trained and evaluated on immediately judgeable proxies such as novelty, clarity, and feasibility. This leaves open whether delayed signals of scientific uptake can be used as feedback for steering models toward research directions with higher expected impact. We study this question using citation-normalized impact as a noisy but scalable proxy for scholarly uptake. We construct a large-scale
dataset from over 100K computer science papers by extracting goal-conditioned idea descriptions and assigning each paper an ordinal, year-normalized citation label. We then train a goal-conditioned reward model to predict citation-impact labels from research goal and idea pairs, and use this reward to align an idea generator through supervised fine-tuning followed by reinforcement learning. To reduce circularity, we evaluate generated ideas with a held-out, reference-grounded protocol that compares model outputs against historical ideas under the same research goal and weights judgments by the reference idea's citation-impact label. Experiments show that our RL-tuned model consistently produces ideas with higher estimated impact than both the base model and supervised fine-tuning baselines. Our findings position scientific impact as a practical, outcome-grounded feedback signal for aligning LLMs in open-ended scientific discovery.
Read the full abstract ▾
Nuclear Theory
Exploring the Parameter Space of pvCD-Bonn Potentials under Constraints from Nuclear Saturation Properties
Ke Nan, Chencan Wang, Jinniu Hu et al. (+2 more)
We explore the parameter space of the pvCD-Bonn B potential within the relativistic Brueckner–Hartree–Fock framework under the constraints from the empirical saturation properties of symmetric
nuclear matter. We first examine the effects of the scalar-
meson and pion coupling constants. Changing the $\sigma$-
meson couplings in the $^{1}S_0$ and $^{3}S_1$–$^{3}D_1$ channels alone cannot reproduce the empirical saturation density and binding energy at the same time, whereas increasing $g_\pi$ moves the saturation point toward the empirical region. However, changing $g_\pi$ alone also affects the deuteron properties and tensor-sensitive observables. We therefore extend the parameter search by varying the $\rho$-
meson tensor coupling $f_\rho/g_\rho$ and readjusting the effective $\sigma$-
meson couplings in the $^{1}S_0$ and $^{3}P_0$ channels. Four representative parameter sets with $g_\pi^2/4\pi=14.7$–15.0 give deuteron properties close to the experimental values and maintain a reasonable description of the main
neutron–
proton phase shifts and differential cross sections at $E_{\rm lab}=50$ and 212 MeV. Their saturation densities lie in the range $0.159$–$0.161~\mathrm{fm}^{-3}$, with saturation energies between $-15.00$ and $-15.38~\mathrm{MeV}$. For the interactions obtained by varying $g_\pi$, the
neutron-
star mass–radius relations show only a weak dependence on the pion coupling, with maximum masses of about $2.24$–$2.28\,M_\odot$ and radii of about $12.3$–$12.6~\mathrm{km}$ at $1.4\,M_\odot$. These results show that including the $\rho$-
meson tensor coupling and partial-wave-dependent $\sigma$-
meson couplings provides a better balance between free-space two-nucleon observables and
nuclear-matter saturation properties than varying the pion coupling alone.
Read the full abstract ▾
All authors Ke Nan, Chencan Wang, Jinniu Hu, Ying Zhang, Hong Shen
Computation and Language
Adaptive Fisher-Whitened Cross-Covariance for Low-Resource Speech Recognition
Asmee Mishra, Mengjie Qian, Brechtje Post et al. (+1 more)
Adapting
multilingual speech foundation models to low-resource
languages remains difficult, especially for
languages that are poorly represented during pre-training. While parameter-efficient fine-tuning (PEFT) reduces the cost of adapting large models, conventional approaches such as LoRA rely on generic low-rank parameterizations and do not explicitly use downstream task information to define the adaptation subspace. To investigate whether task-informed PEFT can better support low-resource ASR, we apply Fisher-Whitened Cross-Covariance Analysis (FCCA) to Whisper and Qwen3-ASR, and introduce two complementary extensions: Asymmetric-Coupled FCCA (AC-FCCA), which exploits structured cross-layer sharing, and Adaptive-Rank FCCA (AR-FCCA), which reallocates adaptation capacity across projection
matrices under a fixed parameter budget. Under controlled
multilingual experiments, we evaluate these approaches on
languages that are poorly represented or unsupported during pre-training alongside well-represented
languages. Standard FCCA is competitive with, and usually outperforms, trainable-parameter-budget-matched LoRA. AR-FCCA provides the most consistent improvement over standard FCCA across both model architectures, with statistically significant gains in several evaluation settings, while retaining the same number of trainable parameters. These results show that task-informed subspace construction can be effective for low-resource
speech adaptation, and that adaptive rank allocation provides a robust way to improve parameter efficiency without increasing model capacity.
Read the full abstract ▾
All authors Asmee Mishra, Mengjie Qian, Brechtje Post, Kate Knill
Optimization and Control
Discrete-time feedback linearization control for nonsmooth constrained optimization
Vito Cerone, Sophie Marie Fosson, Simone Pirrera et al. (+2 more)
We propose a discrete-time feedback linearization control approach to solve nonsmooth linearly constrained composite optimization problems. By using proximal operators, we construct a dynamical system whose
equilibria correspond to the stationary points of the optimization problem. Interpreting the Lagrange multipliers as control inputs, we employ feedback linearization control to steer the obtained dynamical system to a stable
equilibrium. We analyze the convergence of the resulting closed-loop system both in the strongly convex setting and under the Polyak-Lojasiewicz condition. Finally, we illustrate the applicability of the approach through numerical experiments.
Read the full abstract ▾
All authors Vito Cerone, Sophie Marie Fosson, Simone Pirrera, Alice Re, Diego Regruto
Computation and Language
Benchmarking and Domain Adaptation of Automatic Speech Recognition (ASR) for Adolescent Health Communication in Ghanaian Languages
Stephen E. Moore, Akwasi Asare, Mich-Seth Owusu et al. (+3 more)
This paper presents an end-to-end study of automatic
speech recognition (ASR) for adolescent health communication in three Ghanaian
languages (Twi, Dagbani, and Ewe). The work proceeds in three connected stages; First, we
benchmark five ASR systems (three
language-specific Wav2Vec2 models and two
multimodal LLMs, Gemma 3n and Gemma 4) on a general-domain Bible corpus and a Youth Adolescent Sexual and Reproductive Health (ASRH) Domain ASR
dataset, using Character and Word Error Rate (CER, WER). Second, guided by the
benchmark, we perform supervised domain adaptation: although Gemma 4 was the strongest zero-shot candidate, fine-tuning it proved computationally infeasible, so we pivoted to the compact Qwen3-ASR-0.6B, fine-tuned on a large Ghana Bible corpus (~90k samples) and evaluated strictly on held-out human-collected in-domain audio. Fine-tuning reduced WER on every
language, most dramatically for Ewe (WER from 109.3% to 64.8%, a drop of 44.5 pp; CER from 65.1% to 24.9%). Third, we validate the work through KasaHealth, a live voice-first ASRH application deployed in all three
languages, complemented by Senti-Check, a technical evaluation harness. KasaHealth was tested by 50 community respondents and achieved a 100% chat-approval rate, a 72% Good-or-Excellent translation rating, and a 92% would-recommend rate, while surfacing the domain gaps that most constrain real-world use. Across all three stages the evidence converges: for these
languages the binding constraint is validated in-domain data, not model capability or computation.
Read the full abstract ▾
All authors Stephen E. Moore, Akwasi Asare, Mich-Seth Owusu, Paul Azunre, Joel Budu, Lawrence A. Adu-Gyamfi
Complex Variables
Exceptional Sets of Positive Capacity in Gross's Star Theorem
Sina Nadi
In 1918, Gross proved that a regular local inverse of a meromorphic function in the plane can be continued analytically along every ray from its centre except for directions in a set of Lebesgue measure zero. In 1977, Nagasaka asked whether this exceptional set must have logarithmic capacity zero. We prove that there exist a transcendental entire function and a regular local inverse for which the exceptional set contains a compact set of positive logarithmic capacity. In fact, for every $0<s<1$, such a function and inverse can be chosen so that this compact set has positive $s$-dimensional Hausdorff measure.
Read the full abstract ▾
Combinatorics
Generalized Weight Polynomials of Codes through Flats and Orlik-Solomon Algebras of Matroids
Jakob Anderssen and Trygve Johnsen
We present various ways of determining generalized weight
polynomials of a matroid $M$, and we recall how one can find the generalized weight spectra of a linear code, given these
polynomials, for the matroid determined by any generator
matrix of the code. A main goal is to give coding theorists different ways to determine these
polynomials. We describe how one can find the generalized weight
polynomials of any matroid $M$, directly from its
lattice of flats, and we also show how one can find them from the Poincare series (in this case
polynomials) of the associated Orlik-Solomon
algebras of matroids arising as contractions of the flats of $M$. This opens for using information about broken
circuits of the matroids to determine generalized weight
polynomials. We also recall the well-known connection between the Orlik-Solomon
algebra of a matroid, and Whitney numbers obtained by order homology, and use it to show how one can describe weight
polynomials in terms of Whitney numbers from order homology of the matroid and its contraction of flats. We recall briefly a well-known relation between the Orlik-Solomon
algebra of a matroid $M$, which is representable over the complex numbers, and de Rham homology numbers obtained from complements of intersections of hyperplanes in a hyperplane arrangement corresponding to $M$. We also describe a way to find the generalized weight
polynomials of a matroid in terms of
polynomials defined in connection with its
lattice of cyclic flats. In a simple running example we show how one can calculate the generalized weight
polynomials in different ways, including both the methods presented in this paper, and selected methods developed in earlier papers. We also include a less simple example with the projective Reed-Muller code $PR_3(2,2).$
Read the full abstract ▾
Complex Variables
On an Iwaniec–Kovalev–Onninen Conjecture for harmonic quasiconformal annulus mappings
David Kalaj, Jinsong Liu and Jian-feng Zhu
In 2012, Iwaniec, Kovalev and Onninen proposed an upper Nitsche–Grötzsch type estimate for harmonic $K$-quasiconformal homeomorphisms between circular annuli: in normalized form, every such map $h:A(1,s)\to A(1,S)$ should satisfy \[ S \leq \frac{K+1}{2}s-\frac{K-1}{2s}. \] The radial and spiral-radial one-mode models explain why this estimate is natural. We show that this one-mode evidence does not extend to the unrestricted non-radial class. For every $1<s<S$, we construct a non-radial harmonic orientation-preserving diffeomorphism $h:A(1,s)\to A(1,S)$ with \[ \norm{\omega_h}_{\infty}<\frac{S-s}{S-s^{-1}}, \] strictly below the conjectural threshold. Thus, the dilatation lower bound predicted by the
conjecture fails, even for smooth harmonic diffeomorphisms with prescribed circular boundary components. The construction uses a small high-frequency reparametrization of the outer boundary: it lowers the first
Fourier-mode dilatation by order $t^2$, while the compensating high modes are exponentially damped at the inner boundary. We also record structured regimes in which the one-mode estimate survives, including an inner-boundary anti-conformal energy condition, a
Fourier leakage criterion, and a low-frequency spectral stability result. In the minimal-surface interpretation, the upper radial model is helicoidal and vertical-periodic rather than single-valued catenoidal; accordingly, the counterexamples give vertical-periodic minimal annuli with slope below the expected helicoidal threshold, and lead to a non-radial extremal problem for harmonic annulus diffeomorphisms.
Read the full abstract ▾
Materials Science
Anisotropic Magnetic and Transport Properties of RAlGe (R = Y, Gd-Tm, Lu)
Tyler J. Slade, Lin-Lin Wang, Sergey L. Bud'ko et al. (+1 more)
We grew single
crystals of the RAlGe family for R = Y, Gd-Tm, Lu and characterized them with powder X-ray diffraction, temperature dependent specific heat measurements as well as temperature and field dependent resistance and magnetization measurements. These RAlGe materials crystallize in an orthorhombic Cmcm
crystal structure in which the R atoms occupy a position with $m$2$m$ point
symmetry. We find that when R is a moment bearing rare earth atom, the RAlGe materials order antifferomagnetically at $T_{\text{N}}$ ranging from 5 K (TmAlGe) to 39 K (TbAlGe). A second, lower temperature
antiferromagnetic transition is also detected in the range 6-8 K for R = Tb-Ho, but no lower transition is observed (for T above 1.8 K) for R = Er, Tm. The R = Tb-Tm members show evidence for strong
crystal field effects influencing the physical properties, where the
crystal field favors an easy $a$-axis for R = Tb-Ho and an easy $b$-axis for R = Er, Tm. Of these compounds, strongly uniaxial behavior is observed for DyAlGe and TmAlGe. At T = 2 K, rich metamagnetism is observed when the external field is applied along the $a$-axis for R = Tb–Ho, whereas a single metamagnetic transition is observed (up to 70 kOe) with the field along the $b$-axis when R = Er and Tm. All RAlGe materials show metallic transport behavior and have moderate positive magnetoresistance up to 60-150 % at 2 K superimposed on top of metamagnetic features. The field dependent magnetization measurements on non-moment bearing YAlGe show clear de Haas-van Alphen oscillations, indicating the
Fermi surface includes small, high mobility pockets.
Read the full abstract ▾
All authors Tyler J. Slade, Lin-Lin Wang, Sergey L. Bud'ko, Paul C. Canfield
Computational Engineering
Adapting a Large Language Model Crash-Severity Pipeline to Tennessee: Performance Across Sampling Strategies
Abhilasha Saroj, Pranav Govindu, Bharat Sharma et al. (+1 more)
State crash databases differ in structure, coding, and injury-severity distributions, limiting direct reuse of predictive workflows across jurisdictions. This study adapts the SafeTraffic Copilot large
language model (LLM) crash-severity workflow to a three-year Tennessee inventory of 624,392 crashes. Tennessee crash, roadway, vehicle, and person attributes were harmonized and converted into textual
prompts while unavailable values were preserved rather than inferred. Llama 3.1 8B was fine-tuned using low-rank adaptation to classify five injury-severity categories. Random, county-based, and severity-balanced sampling strategies were evaluated using separate in-sample and unseen-test experiments; unseen-test experiments used 70/15/15 training, validation, and test splits. On their respective unseen test sets, random and county-based models achieved weighted F1-scores near 80%, but macro F1 remained below 47% and fatal-crash F1 below 27%, showing that strong aggregate performance can mask weak recognition of rare outcomes. Within its balanced evaluation population, the severity-balanced model achieved weighted and macro F1-scores of 57.7% and a fatalcrash F1-score of 65.9%, yielding more even class-level performance. Because each sampling strategy used a different test subset, crossstrategy differences are descriptive rather than controlled rankings. The results highlight the importance of interpreting LLM crashseverity performance together with sampling design, class balance, class-level metrics, and evaluation-population composition.
Read the full abstract ▾
All authors Abhilasha Saroj, Pranav Govindu, Bharat Sharma, Usman Ahmed
Computation and Language
TimeBraid: Unifying Time Series and Language for Understanding and Forecasting
Xinyue Wang, Jiacheng Pang, Kun Zhou et al. (+7 more)
We present TimeBraid, a series of unified time-series and
language models that align pretrained
language models and pretrained time-series foundation models through interleaved global residual
attention layers. Each model inherits knowledge, instruction following, and
reasoning from one side, continuous-signal perception and zero-shot
forecasting from the other, and fuses the two in a shared representation space where both modalities are understood and generated. We study the design choices that make such unified modeling work: where to align the two representation spaces, how to ground
language in temporal structure, how to balance understanding with generation, and how to keep joint optimization stable. The resulting recipe combines a unified prompting scheme for diverse time-series and text tasks, stabilized joint training, and supervision from 2.2M curated series–text pairs and 4.9M instruction-tuning samples. Across
benchmarks spanning time-series perception, understanding,
reasoning, and both context-aided and unimodal
forecasting, TimeBraid remains competitive with far larger general-purpose models and task-specific counterparts.
Read the full abstract ▾
All authors Xinyue Wang, Jiacheng Pang, Kun Zhou, Kexin Zhang, Defu Cao, Fan Feng, Faisal, Songyao Jin, Yan Liu, Biwei Huang
Analysis of PDEs
Existence and stability of capillary-gravity wave-borne vortices
Gregory Slease, Samuel Walsh and Runzhang Zhong
In this paper, we consider the existence and stability of waves progressing through a finite-depth two-dimensional body of water that carry a
vortex in their bulk. The waves are acted upon by
gravity, and sit below a region of air at constant pressure, with the air–water interface being a free boundary along which capillary effects are incorporated. We consider two classes of
vortices: point
vortices, where formally the vorticity is a Dirac $\delta$, and hollow
vortices, where the
vortex core is a region of constant pressure outside the fluid domain and about which there is a nonzero circulation. First, we prove that steady small-amplitude wave-borne point
vortices exist for any subcritical Froude number and supercritical Bond number, and then show they are conditionally orbitally stable. Second, through a
vortex desingularization argument, we construct capillary-
gravity wave-borne hollow
vortices. Notably, the wave speed is $O(1)$ for both these families.
Read the full abstract ▾
Optimization and Control
Budget-Constrained Graph Augmentation for Robust Network Design via Kirchhoff Index Minimization
Omkar Bhoite, V Sateeshkrishna Dhuli, Stefan Werner et al. (+1 more)
Enhancing the robustness of deployed networks against failures and disruptions is critical for reliable operation. This requires deciding which new links to install and how strongly to weight them under limited resources. We study this problem through the Kirchhoff index, or total effective resistance, a spectral measure of global connectivity. The resulting augmentation problem couples discrete candidate-edge selection with continuous weight allocation under heterogeneous per-unit deployment costs, a total budget, and an exact-cardinality constraint. For a fixed weighted base graph, this yields a mixed-integer formulation and a semidefinite relaxation whose optimum lower-bounds the mixed-integer optimum. We cast the relaxation as a cone program and solve it numerically using a homogeneous self-dual
embedding and first-order operator splitting. Feasible discrete designs are recovered through rounding-and-repair procedures and assessed by a posteriori gap estimates relative to the numerical semidefinite program (SDP)
benchmark. As a scalable alternative, we develop an exact-$k$, budget-feasible greedy heuristic built on rank-one
Laplacian updates and biharmonic-distance caching, and interpret its progress through a Bellman value-to-go
benchmark with a conservative spectral lower bound on the local policy ratio. Experiments on synthetic and real infrastructure networks across graph sizes, budgets, weight distributions, and cost regimes show that, under fixed budgets, distance-proportional costs limit the achievable resistance reduction and shift installed conductance toward shorter links relative to uniform per-unit costs.
Read the full abstract ▾
All authors Omkar Bhoite, V Sateeshkrishna Dhuli, Stefan Werner, Kimmo Kansanen
Materials Science
Unified description of exciton, phonon and plasmon dispersions in 2D materials from an optical conductivity approximation
Alberto Guandalini, Andrea Ferretti, Daniele Varsano et al. (+2 more)
The modified Coulomb interaction in two-dimensional (2D) materials gives rise to unique, non-analytic low-momentum dispersions of
excitons,
phonons, and plasmons. Here, we describe the 2D dispersion of longitudinal excitations by introducing the optical conductivity approximation (OCA), a unified framework that evaluates the finite-q longitudinal response probed by electron energy-loss spectroscopy (q-EELS) by only using the q = 0 optical conductivity. By doing so, the approach demonstrates that the linear dispersion of all the above mentioned excitations is mainly due to the form of the 2D macroscopic Coulomb interaction rather than to the dispersion of the underlying band structure. Applying this approach to hexagonal boron nitride and
graphene we accurately reproduce the energy and intensity dispersions of the hBN longitudinal-optical
phonon and low-energy bright
excitons, along with the
graphene pi-plasmon. Our work also provides an efficient computational scheme to describe low momentum EELS data without the need to calculate the response functions at dense moment
Read the full abstract ▾
All authors Alberto Guandalini, Andrea Ferretti, Daniele Varsano, Paolo Barone, Francesco Mauri
Computer Vision
OREO: Fidelity Alignment in 3D Generation via On-the-fly Rendering-Editing Optimization
Zhiyuan Ma, Wenbo Hu, Wang Zhao et al. (+3 more)
Despite recent advancements in 3D generation, models often struggle to produce assets with high visual fidelity. To bridge this gap, we propose OREO, an
alignment framework that enhances the realism of 3D generators by leveraging rich 2D
diffusion priors. Instead of relying on static
datasets, OREO establishes a dynamic optimization loop that produces on-the-fly edited renderings as 2D pseudo-targets. At its core, we introduce Reinforced Editing, which utilizes a 2D model to refine rendered views of the 3D output, enhancing their overall visual fidelity while preserving the underlying geometry, viewpoint, and content. These refined views serve as high-quality supervision targets, enabling the 3D generator to learn from its own generated samples and progressively improve its visual quality. Experiments demonstrate that OREO effectively improves upon pre-trained baselines, producing 3D assets with enhanced visual realism.
Read the full abstract ▾
All authors Zhiyuan Ma, Wenbo Hu, Wang Zhao, Pengfei Wang, Ying Shan, Lei Zhang
High Energy Astrophysical Phenomena
Observational signatures of warped accretion flows in tidal disruption events
Andrew Mummery and Eliot Quataert
We analyze the possible observational signatures and
evolution of global warps in the
accretion flows produced in
tidal disruption events. These warps are driven by Lense-Thirring torques resulting from the misalignment of the disk's rotation axis and the black hole's
spin axis. In a previous paper we argued that these global warps should be generically present in TDE
accretion flows as they enter a late-time thin disk phase of
evolution. We isolate three possible observational signatures of warps: (i) anomalous (reddened) late time optical colors, resulting from irradiative heating of the warped outer disk by the inner flow, (ii) delayed X-ray rises, resulting from the inner disk being slowly revealed to a distant observer as the warped flow gradually aligns with the
spin axis, and (iii) reverberation of turbulent X-ray
flares into optical/UV frequencies as the inner disk heats the outer disk during a flaring state. We present preliminary observational evidence that all three effects may have already been observed in TDE systems. Observations of TDEs may well offer the cleanest astronomical signatures of the strong-
gravity effects which cause disk warping, and provide unique constraints on the poorly understood
evolution of warped
accretion disks.
Read the full abstract ▾
Quantum Gases
Microscopic theory of the collective optical response of dilute atomic clouds at finite temperature
Tanja Schoger, Romain Pierrat and Nicolas Cherroret
At low temperature, collective light-induced dipole-dipole interactions are known to strongly reshape the optical response of atomic clouds, revealing the breakdown of the independent-scattering picture as the density increases. While thermal motion is generally believed to progressively suppress these interactions, the microscopic mechanisms behind this suppression remain largely unexplored. Here, we develop a microscopic theory of the optical response of a dilute atomic gas at finite temperature, explicitly accounting for atomic ballistic motion during the recurrent-scattering events associated with the collective corrections. Using a diagrammatic approach for scalar light, we derive the collective contribution to the optical permittivity and characterize its behavior across the full temperature range. We find that thermal corrections scale as $\sim T$ at low temperature, while recurrent-scattering contributions are suppressed as $\sim T^{-3/2}$ at high temperature. Our predictions are confirmed by extensive coupled-dipole simulations that explicitly account for atomic motion. These simulations also reveal the inaccuracy of the modified frozen-dipole approximation commonly used to treat thermal effects. Finally, extending our theory to vector light, we provide a complete picture of how the resonance shift continuously evolves from the collective Lamb shift at low temperatures to the classical Lorentz-Lorenz shift at high temperatures.
Read the full abstract ▾
Computer Vision
Sheekar Banerjee, Md. Srabon Chowdhury, Md. Mahbub Hasan Akash et al. (+1 more)
Accurate
brain tumor
segmentation from
Magnetic Resonance Imaging is essential for diagnosis, treatment planning, and
surgical guidance. Although Convolutional Neural Networks, particularly UNet, have achieved significant success in medical image
segmentation, they often struggle to capture the long-range spatial dependencies required to model tumors with irregular shapes and complex boundaries. This paper proposes a lightweight
Vision Transformer UNet that combines the hierarchical feature extraction capability of UNet with the global context modeling of
Vision Transformers. The proposed architecture incorporates a compact ViT bottleneck within a U-Net encoder-decoder framework, enabling effective learning of both local and global features while maintaining computational efficiency with only 2.6 million trainable parameters. The model was evaluated on the TCGA LGG MRI
Segmentation dataset, achieving a mean Intersection over Union of 0.8100 and a Dice score of 0.8446, outperforming the baseline UNet by 3.75% and 3.15%, respectively. Extensive quantitative and qualitative analyses, including confusion
matrix evaluation, precision recall curves, per-image performance distribution, and tumor size dependency analysis, demonstrate the effectiveness and robustness of the proposed method for
brain tumor
segmentation.
Read the full abstract ▾
All authors Sheekar Banerjee, Md. Srabon Chowdhury, Md. Mahbub Hasan Akash, Ishtiak Al Mamoon
Combinatorics
Congruence Classes in $\mathbb{F}_p^2$: Sharp Results via an Energy Approach
Thang Pham, Le Quang Ham, Steven Senger et al. (+2 more)
Let $T_k(E)$ denote the set of congruence classes of ordered $k$-tuples of pairwise distinct points of $E$. Let $p\equiv3\pmod4$ be
prime. For $E\subset\mathbb{F}_p^2$ with $3\leq|E|\leq p^{3/4}$, we prove that $|T_3(E)|\gg|E|^{11/6}$; for $4\leq|E|\leq p^{3/4}$, we prove that $|T_4(E)|\gg|E|^3/\log|E|$. For every fixed $k\geq5$ and $k\leq|E|\leq p^{3(k-2)/(3k-4)}$, we prove that $|T_k(E)|\gg_k|E|^{k-1}$. The proofs proceed by bounding the moments of the overlap function of rigid motions. The main inputs are an exact identity involving the distance energy and an incidence bound obtained by viewing rigid motions as lines over $\mathbb{F}_p(i)$. The bound for $k=4$ is sharp up to a logarithmic factor, while the bounds for $k\geq5$ are sharp up to constants.
Read the full abstract ▾
All authors Thang Pham, Le Quang Ham, Steven Senger, Dung The Tran, Boqing Xue
Earth and Planetary Astrophysics
VLTI/MATISSE observations of the hot dust around $\beta$ Pictoris: Observing faint objects with MATISSE
P. Priolet, J.-C. Augereau, J. Milli et al. (+32 more)
Context: Hot exozodiacal
dust has been detected around approximately 20% of nearby main-sequence
stars, indicating that it is a common feature of extrasolar planetary systems. However, its spatial distribution and grain properties remain poorly understood. Aims: We aim to show that the spatial distribution of hot
dust around beta Pictoris, an iconic planetary system with two known
planets and a population of exocomets, can be constrained through a detailed analysis of optical long-baseline interferometry data. Methods: We obtained L-band observations at approximately 3.5 microns of beta Pictoris with MATISSE at the VLTI. We developed a method to assess correlations in the visibility measurements by estimating the full data covariance
matrix, and modeled the circumstellar emission with
geometric brightness distributions. Results: We detect circumstellar emission around beta Pictoris consistent with hot exozodiacal
dust. Modeling the emission with and without accounting for data correlations demonstrates their significant impact on the inferred
dust properties. The best-fit model reveals a compact (approximately 0.05 au), highly inclined
dust structure near the sublimation radius, misaligned by approximately 60 degrees on the sky with respect to the extended edge-on outer disk. It contributes approximately 4%-7.5% of the total flux at 3.4 microns and is composed of submicron-sized grains. The data are also compatible with an additional
dust component near 1 au. Conclusions: For beta Pictoris, the two-decade-old paradigm placing hot exozodiacal
dust near the sublimation radius and consisting of submicron-sized grains finds observational support in the detailed analysis of MATISSE data. The misalignment with the known debris disk further supports a link to exocometary activity.
Read the full abstract ▾
All authors P. Priolet, J.-C. Augereau, J. Milli, A. Matter, B. Lopez, H. Beust, J. Varga, P. Boley, W.-C. Danchi, L. Foteini, T. Henning, M. Houllé, M. Letessier, F. Millour, J. Scigliuto, G. Weigelt, S. Wolf, O. Absil, J.-P. Berger, Y.-I. Bouarour, G. Bourdarot, D. Defrère, C. Desgrange, J.-B. Le Bouquin, D. Iglesias, T. A. Stuber, P. Berio, F. Bettonvil, P. Cruzalèbes, M. Heininger, J. W. Isbell, S. Lagarde, A. Meilland, R. Petrov, S. Robbe-Dubois
Astrophysics of Galaxies
How Magnetic Fields Regulate Cooling and Mixing in Turbulent Radiative Mixing Layers
Rajsekhar Mohapatra, Lachlan Lancaster, Drummond Fielding et al. (+2 more)
Turbulent radiative mixing layers (TRMLs) are expected wherever hot and cold gas move past one another, including in the
solar corona,
galactic winds, and cold filaments in
galaxy clusters. These environments are often magnetized, but
magnetic effects on mixing and cooling remain less well understood than in the hydrodynamic (HD) case. We present magnetohydrodynamic (MHD) simulations of TRMLs at resolutions up to $1024 \times 2048^2$, spanning fields aligned with and transverse to the shear, and polarity-reversing configurations in which oppositely directed fields form current sheets at the interface. Even initially weak hot-phase fields, with $\mathcal{M}_{\rm A,shear}\equiv v_{\rm shear}/v_{\rm A}\sim14$, reduce the mass and enthalpy flux and radiative cooling rate by up to an order of magnitude relative to HD.
Magnetic tension weakens turbulent motions, reducing both the
diffusion of hot gas into the layer and the folding of the cooling surface. Transverse fields suppress cooling somewhat more strongly than shear-aligned ones, although a transverse field exerts no tension against the initial linear instability. Polarity reversal changes the morphology of the cooling gas without restoring HD-like mixing. The dependence on the Damköhler number is similar to the HD case, but the degree of suppression is dependent on the initial field orientation. The net cooling rate appears resolution-independent in HD but declines with resolution in MHD, and has not converged, so our suppression factors are lower limits.
Magnetic fields therefore strongly regulate cooling and mixing in multiphase gas, and quantitative predictions require careful treatment of transport processes.
Read the full abstract ▾
All authors Rajsekhar Mohapatra, Lachlan Lancaster, Drummond Fielding, Eliot Quataert, Greg L. Bryan
Artificial Intelligence
Huseyin Cavus, Sebin Sabu, Joshua Spear et al. (+2 more)
Interpretable machine learning for Large
Language Models (LLMs) increasingly relies on sparse probing methods that identify small sets of
neurons claimed to detect and causally influence behaviors such as factuality recall, safety
alignment, and
hallucination. These claims have important implications for model
auditing and behavioral steering, yet they are rarely tested against known failure modes of $L_1$-regularized probing in correlated, high-dimensional feature spaces. We propose a five-step diagnostic protocol covering feature correlation, bootstrap stability, sparse versus dense ranking disagreement, intervention baselines, and cross-
dataset evaluation as a minimum standard for sparse-
neuron localization claims. We investigate prior work using our proposed approach, specifically on H-
neurons using open-source LLMs across TriviaQA, BioASQ, and NQ-Open
datasets. Our results demonstrate detection replicates across both models and
datasets, and exceeds the original reported AUROC gaps for TriviaQA and BioASQ
datasets. Gemma 3 4B consistently outperforms MedGemma 4B on matched
datasets, with AUROC gaps of +0.311 versus +0.235 on TriviaQA, +0.474 versus +0.455 on BioASQ, and +0.128 versus +0.112 on NQ-Open respectively. Causal validation at $n = 500$ with five random seeds shows statistically significant effects beyond random same-layer baselines. At the same time, the diagnostic results indicate that the selected
neurons are not uniquely localized. Across the three Gemma 3 4B settings, 19 of 22 selected H-
Neurons have Pearson $|r| > 0.7$ with other features, bootstrap selections show only moderate stability, and sparse and dense rankings overlap only weakly. Our findings show that sparse predictive structure can coexist with non-unique
neuron selection. Routine diagnostic validation is necessary to distinguish detection claims from localization claims in mechanistic interpretability.
Read the full abstract ▾
All authors Huseyin Cavus, Sebin Sabu, Joshua Spear, Jaskaran Singh Kawatra, Pavithra Rajendran
Sound
Joint Analysis of Latent Dimensionality and Frame Rate in Continuous Audio Encoders
Kyudan Jung, Sehyun Lee, Song-ha Jo et al. (+2 more)
Continuous audio encoders compress audio along feature and time axes through latent width and frame rate, but their joint effect on downstream performance remains unclear. We train sixteen encoders spanning four widths and four frame rates, with downstream adapters and probes, using matched training protocols. Despite generally improved reconstruction at larger widths, automatic
speech recognition (ASR) and spoken question answering (SQA) favor moderate widths at higher rates, with the best observed widths shifting toward larger values under stronger temporal compression. Frozen-model PCA interventions reveal distinct reconstruction and recognition sensitivities: removing the trailing half of the components substantially degrades ASR in high-rate 512-dimensional encoders with comparatively small reconstruction penalties, whereas 1024-dimensional encoders largely preserve both. Yet the projected 1024-dimensional model underperforms unmodified narrower models on ASR at 12.5Hz. These findings identify a width–rate interaction in downstream utility and suggest that how representations are organized during training matters beyond reconstruction fidelity and compressibility.
Read the full abstract ▾
All authors Kyudan Jung, Sehyun Lee, Song-ha Jo, Jaegul Choo, Sanghyuk Choi