Statistical Mechanics
Static versus dynamic universality in the site-diluted kagome Ising model
Alexandros Vasilopoulos, Zeynep Demir Vatansever, Erol Vatansever et al. (+2 more)
We study the site-diluted Ising model on the kagome
lattice using the Wolff single-cluster
algorithm, focusing on both
equilibrium and dynamic critical properties. The
equilibrium critical exponents $\nu = 1$ and $\gamma/\nu = 7/4$ retain their exact pure Ising values for all
spin concentrations $p$ considered, consistent with the marginal irrelevance of disorder at $\alpha = 0$, with the specific heat crossing over from logarithmic to double-logarithmic growth upon dilution. These static results are shared between the kagome and square
lattices. On the dynamic side, $z$ decreases upon dilution on both
lattice, opposite to what is observed under local dynamics, with evidence of saturation to a $p$-independent value at strong dilution, suggesting a diluted dynamic fixed point. While $z$ is consistent between the two
lattices in the pure case, it differs in the diluted regime, an effect we attribute to the two
lattices sitting at different positions along the renormalization-group flow between the Ising and percolation fixed points, rather than to a true asymptotic breakdown of dynamic universality.
Read the full abstract ▾
All authors Alexandros Vasilopoulos, Zeynep Demir Vatansever, Erol Vatansever, Gerard T. Barkema, Nikolaos G. Fytas
Biomolecules
RNA Design via Conditioned Flow Matching and Finite-Policy Reinforcement Learning
Zefeng Lin, Xianyong Fang, Tianfan Fu et al. (+1 more)
RNA design aims to identify sequences that fold into specified secondary structures. Existing methods formulate the task as target-specific search or conditional generation. However, natural RNA
evolution proceeds through sequence variation and selection, with compensatory substitutions, whereas these methods do not explicitly model this process. To address this limitation, we propose a two-stage framework comprising RNA Inverse-Folding Flow (RNA-IFlow) and RNA-IFlow-RL. RNA-IFlow uses structure-conditioned Dirichlet Flow Matching to model coordinated variation across the sequence, while RNA-IFlow-RL maps the learned flow to a pairing-preserving finite policy and refines it with thermodynamic feedback. Our framework achieves leading performance on multiple
benchmarks, reaching 85.19% Pass@1 on Rfam-27. Further analyses reveal thermodynamic gains, policy dynamics, and robustness across settings. Our work couples coordinated variation with thermodynamic selection, offering a novel paradigm for RNA design.
Read the full abstract ▾
All authors Zefeng Lin, Xianyong Fang, Tianfan Fu, Xiaohua Xu
Algebraic Topology
Stable rational cohomology splitting of moduli of branched covers of curves
Andrea Bianchi
We consider the
moduli stack $\mathcal{H}_{g,d}$ parametrising maps $\theta\colon C\to L$ of degree $d$ between complex curves of genera $g$ and $0$. We provide a splitting of the stable rational
cohomology of $\mathcal{H}_{g,d}$, for $g\to\infty$, in terms of degrees of intermediate covers, and identify each direct summand in terms of factorisation homology. For $2\le d\le 5$, we further interpret the stable rational
cohomology in terms of configuration spaces with labels in partial abelian monoids; we show in particular the existence of stable odd-dimensional
cohomology classes for $d=4,5$.
Read the full abstract ▾
Machine Learning
Architecture Alignment With Sparse Priors in Tabular Foundation Models
Tianqi Zhao, Tianyi Zhuang, Shuo Duan et al. (+3 more)
Tabular foundation models (TFMs) are increasingly popular because they deliver strong predictions on new
datasets through in-context learning, without task-specific training or extensive tuning. Yet released TFMs differ simultaneously in their pretraining priors, architectures, and objectives, obscuring their respective inductive biases. We therefore examine one concrete capability: irrelevant-feature suppression. Across synthetic tasks and real-world
datasets, adding null features causes substantially greater predictive degradation in the row-
token model TabDPT, whereas the
cell-
token alternating-axis model TabPFN v2 and other TFMs remain comparatively stable. This gap motivates us to ask whether architecture contributes to irrelevant-feature suppression. Because released TFMs remain confounded by other design choices, we train streamlined row-
token and alternating-axis
transformers under identical sparse-to-dense linear priors. Exact Bayes analysis shows that sparse prediction requires context-dependent feature gating, whereas the dense endpoint requires only uniform feature weighting. Consistent with this distinction, the alternating-axis model is substantially closer to the
Bayesian optimal predictor on sparse tasks, while the architecture gap becomes negligible on dense tasks; almost all of the sparse gap arises from linear coefficient-estimation error. Finally, in both the controlled model and frozen TabPFN v2, we examine the effect of interventions on the feature-
attention outputs on the linear coefficients, finding evidence of task-dependent selective routing of computation through feature-indexed pathways. Together, these results support architecture-prior
alignment: preserving an addressable feature axis provides an inductive bias for task-adaptive relevance inference. Code is available at https://github.com/Tianqi-Zhao/ArchitecturePriorTFMs.
Read the full abstract ▾
All authors Tianqi Zhao, Tianyi Zhuang, Shuo Duan, Guanyang Wang, Yan Shuo Tan, Qiong Zhang
Computer Vision
Less Supervision, Better Generalization: Weakly Supervised Fake Region Localization in Diffusion-Edited Images
Junhee Lee, Donghyeon Jeon, Taeoh Kim et al. (+2 more)
Localizing AI-edited regions is essential for interpretable forensic analysis, but remains challenging due to subtle and spatially distributed artifacts that are misaligned with semantic or object boundaries. Existing approaches rely on pixel-level supervision from controlled editing pipelines, which is difficult to scale and can introduce misleading signals: artifacts frequently extend beyond annotated regions, while out-of-mask pixels are treated as authentic. This limits models' ability to capture transferable evidence and generalize across generators and
datasets. To address these issues, we propose ReGFLoW, a Reconstruction-Guided Fake Localization framework under Weak supervision, which is the first weakly supervised approach for
diffusion-edited fake region localization. ReGFLoW requires only real/fake labels at the image-level and uses
diffusion reconstruction errors as dense spatial guidance to inject them into both feature and score spaces. Furthermore, by artifact-centric multiple instance learning, ReGFLoW utilizes localized
diffusion evidence without relying on semantic-affinity or boundary-based pseudo-mask priors. Extensive experiments show competitive cross-generator localization, while ReGFLoW outperforms all evaluated fully supervised baselines when evaluation includes both partially edited and fully synthetic images and in cross-
dataset tests, without target-domain adaptation.
Read the full abstract ▾
All authors Junhee Lee, Donghyeon Jeon, Taeoh Kim, Beomyoung Kim, MyeongAh Cho
Machine Learning
What You Observe Determines How You Identify Causal Effects: Evaluating Causal Models across Observational Views
Heejin Jung, Gyeongdeok Seo, Hoyoon Byun et al. (+2 more)
Causal foundation models (CFMs) pre-trained on data generated from various structural causal models (SCMs) have been proposed for estimating causal effects from observational data. However, differences in pre-training environments and evaluation protocols make it difficult to assess how their performance depends on the information available for causal identification. To enable controlled comparisons, we introduce CausalIDView, a multi-view
benchmark that holds fixed SCM realization and target estimand while varying only the observational view available to the
estimator. Each observational view corresponds to a distinct identification regime under the
benchmark's maintained causal assumptions. Across these matched views, no CFM consistently performs best and model rankings vary substantially. Under controlled structural changes, CFMs exhibit model-specific failures to maintain stable estimates when true effects are unchanged and to track genuine effect changes. We also examine whether combining explicit identification with strong predictive estimation is effective. A modular approach that pairs a predictive tabular foundation model with regime-specific identification procedures is competitive with CFMs and outperforms several of them. These findings motivate cross-regime comparisons to assess the empirical value of CFMs.
Read the full abstract ▾
All authors Heejin Jung, Gyeongdeok Seo, Hoyoon Byun, Joseph Lee, Kyungwoo Song
General Relativity and Quantum Cosmology
Higher-curvature gravity in Weyl geometry: No-ghost and no-tachyon conditions
Tomoya Tachinami
Weyl geometry, in which the length scale can be chosen independently at each point of spacetime, provides a natural framework for extending general relativity through Weyl invariance, that is, local scale
symmetry. We construct a gravitational theory whose Lagrangian is an arbitrary function of the scalar
curvature in Weyl geometry and establish the conditions under which the theory is free of ghosts and tachyons around maximally symmetric
vacua. For an arbitrary function, Weyl invariance requires a compensating scalar field in addition to the metric and the Weyl vector. By linearizing the action with an auxiliary field, we show that the Weyl vector couples to the scalar sector only through a single combination of the scalar fields. Consequently, the Weyl vector, shifted by the logarithmic gradient of this combination, becomes a Weyl-invariant massive vector field without gauge fixing. In the Einstein frame, given that the graviton is ghost-free, the no-tachyon condition for the Weyl vector, together with the no-ghost condition for the scalaron, the scalar mode associated with the higher-
curvature terms, reduces to the positivity of the kinetic coefficient of the compensating scalar. When this coefficient vanishes, the scalaron does not propagate, leaving only the graviton and the massive vector. As an example, we show that for a Lagrangian containing constant, linear, and quadratic
curvature terms, both a massive vector and a massive scalaron are necessarily present throughout the stability region, whereas without the linear term the scalaron becomes massless.
Read the full abstract ▾
Cryptography and Security
SKILLLITE: Evidence-Guided Malicious Skill Auditing with Compact LLMs
Haoran Ou, Gelei Deng, Xuanye Zhang et al. (+3 more)
As LLM-based
agents perform increasingly complex tasks,
Agent Skills have emerged as a flexible mechanism for extending their capabilities. An
Agent Skill packages task-specific instructions with executable components and auxiliary resources to provide specialized functionalities. However, the growing adoption of third-party Skills introduces a new supply-chain attack surface. Malicious Skills can embed harmful behaviors that abuse
agent privileges and compromise the
agent execution environment or accessible resources. Although recent LLM-based malicious Skill
auditing approaches have achieved promising performance, they often rely on capable commercial LLMs. How to achieve effective
auditing with compact, locally deployable LLMs in security-sensitive and resource-constrained settings remains largely unexplored. Our investigation reveals that compact LLMs struggle to identify malicious behaviors hidden in complex Skill packages. This difficulty arises from both the implicit nature of such behaviors and the limited
reasoning capacity of compact LLMs. To address these challenges, we propose SKILLLITE, an evidence-guided agentic framework for malicious Skill detection. SKILLLITE effectively extracts security-relevant behaviors and infers the intended functionality from complex Skill packages. It then employs a compact LLM to assess the maliciousness of the Skill based on the observed behaviors and their functional context. Experiments show that SKILLLITE improves malicious Skill detection across different compact LLM backbones and outperforms existing representative
auditing baselines. Its effectiveness generalizes to behaviorally confirmed in-the-wild malicious Skills. Meanwhile, SKILLLITE maintains a low inference latency, supporting its practical deployment.
Read the full abstract ▾
All authors Haoran Ou, Gelei Deng, Xuanye Zhang, Wenbo Guo, Tianwei Zhang, Kwok-Yan Lam
Robotics
NIDAR: NIR-Guided Intrinsic Decomposition for Scalable Scene-Agnostic LiDAR Intensity Reconstruction
Junjie Zhang, Jie Yin, Kefei Qian et al. (+5 more)
LiDAR return intensity provides complementary surface-response cues for robotic perception and state estimation, yet many simulation pipelines omit it or reproduce it using reconstruction methods that require real intensity supervision and per-scene optimization. These requirements increase data-collection and fitting costs and limit reuse across simulated scenes. We present NIDAR, a feed-forward framework that synthesizes dense intensity-like observations from RGB appearance and simulator geometry. NIDAR combines pretrained pseudo-NIR translation, hierarchical intrinsic decomposition, geometry-aware modulation, and source-domain distribution calibration to transfer reflectance-related image cues to simulated point clouds. Its learned components are trained offline using Waymo data; their weights and calibration remain fixed during evaluation on Waymo and nuScenes. Deployment therefore requires neither target-scene intensity labels nor target-scene gradient-based fitting. The reported comparisons show competitive pixel-wise accuracy and favorable structural and perceptual fidelity against the evaluated reconstruction baselines. A controlled pseudo-NIR-versus-RGB diagnostic further shows that the pseudo-NIR prior is most beneficial when used through the paper-aligned reflectance-and-remapping route, rather than as a simple direct intensity regressor. We further integrate NIDAR with Unreal Engine 5, Isaac Sim, and a generative
LiDAR pipeline. Two intensity-aware SLAM systems evaluated in two simulated indoor scenes suggest potential downstream utility, but do not constitute real-
robot validation. NIDAR therefore offers a scalable intensity-synthesis interface for the evaluated settings; cross-wavelength, camera-configuration, embedded, and real-sensor validation remain future work.
Read the full abstract ▾
All authors Junjie Zhang, Jie Yin, Kefei Qian, Jie Li, Mengpei Jia, Yajuan Dun, Wenbo Chu, Guofa Li
Number Theory
The Index of Composition of polynomials and its applications
Surender Kumar and Sumandeep Kaur
The index of a given monic irreducible
polynomial $f(x)\in\mathbb Z[x]$ having a root $\theta $ is the index of $\mathbb Z[\theta]$ in the ring of algebraic integers of $\mathbb Q(\theta).$ Further, $f(x)$ is monogenic if its index is $1$. In this article, we prove that the index of $f(x)$ always divides the index of $f(g(x)),$ where $f(x), g(x)\in \mathbb Z[x]$. Additionally, we give the precise power of the index of $f(x)$ dividing the index of $f(g(x))$. Also, we provide a necessary condition for the monogenity of $f(g(x))$. As an application of our results, we prove that the $n$-fold composition $f^n(x)=(f\circ f\cdots\circ f)(x)$ of a
polynomial $f(x)$ is monogenic for all $n\in\mathbb{N}$ if and only if the sequence $\langle I_n\rangle$ is convergent, where $I_n$ is the index of $f^n(x)$. Moreover, we show that for a given $n\in\mathbb{N}$, if $I_n$ is $m^{\text{th}}$ power free integer, then the $n$-tower $\mathbb Q\subseteq K_1 \subseteq\cdots \subseteq K_{n-1} \subseteq K_n,$ defined by $n$ iterates of a
polynomial $f(x)\in\mathbb Z[x]$ contains at least $\max\{\lfloor n-\log_{\deg f}m \rfloor,0\}$ monogenic number fields.
Read the full abstract ▾
Nuclear Theory
Wigner and mirror correlations in nuclear mass predictions with kernel ridge regression
X. H. Wu
Wigner- and mirror-correlated kernel ridge regression (WKRR) and its odd-even extension WKRRoe are developed to improve
nuclear mass predictions. The Wigner and mirror correlations are encoded entirely in the kernel functions without introducing more weight parameters. For experimentally known
nuclear masses, WKRRoe achieves a leave-one-out root-mean-square deviation of 98.0 keV, which falls below the often-quoted 100-keV scale discussed in connection with chaos-related limits. The gain is mainly concentrated in light
nuclei near $N=Z$ and is especially large for mirror pairs. Further tests show that WKRRoe can potentially improve predictions for experimentally unknown
nuclei near $N=Z$, particularly when their mirror partners have already been measured.
Read the full abstract ▾
Computer Vision
S4VY: Segment Anything in Feed-Forward 4D Visual Geometry
Jingdong Zhang, Xin Li, Jan Kautz et al. (+2 more)
Accurate instance
segmentation in dynamic scenes is important for downstream applications such as robotics and autonomous driving. Existing Segment Anything models operate primarily on 2D image or
video masks and preserve identity through sequential
memory, while promptable 4D instance
segmentation built upon feed-forward visual geometry remains underexplored. We introduce S4VY, a Segment Anything model built on feed-forward 4D visual geometry. From a set of RGB observations, S4VY transforms shared visual-
geometric features into an exhaustive set of class-agnostic 4D instance masks through a space-time query decoder, with each persistent object query binding one entity across all observations. This representation supports
prompt-independent
segmentation as well as point- and box- conditioned selection, without requiring a seed mask or temporal ordering. We further develop an agentic harness for natural-
language grounding in the large observation space of a 4D scene. Active tree search identifies relevant frames without scanning every fixed window; a dual-stream grounder combines fine-grained VLM visual priors with geometry-consistent instance features through complementary bounding-box prediction and object-query matching; and an independent critic selects the final 4D instance mask from their predictions. Extensive experiments demonstrate state-of-the-art 4D instance
segmentation and strong
language-guided grounding performance under a unified evaluation spanning static and dynamic scenes.
Read the full abstract ▾
All authors Jingdong Zhang, Xin Li, Jan Kautz, Wenping Wang, Chris Choy
Quantum Physics
Closing a single-copy gap to the Holevo bound with two-copy measurements
Changhao Li
The Holevo Cramér-Rao bound sets the asymptotically attainable precision of
quantum multiparameter estimation, but its attainability with finite
quantum state copies remains unresolved in general. While a gap between the single-copy optimum and the Holevo bound persists at any finite copy number for two-parameter estimation models, the higher-parameter case has remained open. Here we disprove universal finite-copy gap persistence beyond two parameters by constructing a three-parameter rank-two model whose Holevo bound is unattainable with one copy but attained exactly with two. We further prove finite-copy persistence for full-rank models with any number of parameters. We then experimentally implement the two-copy collective measurement for the constructed model using superconducting
qubits. We observe a two-copy estimation error below the single-copy and separate-measurement bounds determined from state
tomography. By introducing controlled depolarization, we show that this precision advantage persists over a finite noise range. Our results connect the attainability of ultimate
quantum precision to the structure of the multiparameter estimation model and the resources available for collective measurements.
Read the full abstract ▾
Machine Learning
Seeing Time: Visual-Temporal Representation Learning for Interpretable Time Series Clustering
Zheng Zhu, Zexi Tan, Yuming Deng et al. (+1 more)
Multivariate Time Series (MTS) clustering is an important tool in temporal data mining, aiming to discover latent group structures from complex observations without supervision. Although existing deep clustering methods can learn discriminative temporal representations, the resulting latent clusters are often difficult to relate back to waveform characteristics that practitioners can directly inspect and compare, limiting their ability to assess whether the discovered patterns reflect meaningful temporal behaviors. This paper, therefore, proposes WAVE (Waveform Aligned Visual-temporal
Embedding), which treats time series and their deterministically rendered waveform plots as complementary views of the same observations. To produce discriminative representations whose cluster structures can be traced to observable waveform characteristics, WAVE aligns and integrates fine-grained temporal variations with holistic visual patterns, while associating each discovered cluster with its centroid-nearest authentic sample. Accordingly, interpretability in this work specifically refers to waveform-level traceability rather than a general explanation of model decisions. Extensive evaluations across 10 real-world public
datasets show that WAVE achieves the highest macro-averaged clustering performance and the best average rank among the compared methods, while qualitative case studies illustrate how the discovered clusters can be inspected through authentic waveform records. The source code is available at https://github.com/Zheng-Zhu1/WAVE.
Read the full abstract ▾
All authors Zheng Zhu, Zexi Tan, Yuming Deng, Yiqun Zhang
Robotics
PreferenceFlow: Test-Time Guidance of Flow-Matching Robot Policies from Human Interventions
Yiqi Tang, Diyuan Shi, Runze Li et al. (+1 more)
Flow-matching policies can represent complex
robot behaviors but remain susceptible to local errors under distribution shift at deployment. Many reinforcement learning approaches to policy improvement require reward signals that are difficult to specify or obtain in real-world manipulation. We present PreferenceFlow, a framework for improving a pretrained flow policy at test time without environment rewards or updates to the base policy. Human intervention chunks are paired with
robot chunks generated from the same initial conditioning state to train a preference model. During infer- ence, we adopt the QGF sampling update, replacing its value gradient with the preference gradient evaluated at an estimated clean action. A gradient-cap loss penalizes excessive gradients on intervention pairs, while a zero-gradient loss discourages guidance near actions from expert demonstrations. On four real-world precision insertion tasks with a Franka
robot, Pref- erenceFlow achieves a mean success rate of 90.5%, compared with 69% for the frozen policy. Ablations support the roles of gradient regularization and correctly ordered preference labels in the evaluated settings. These results demonstrate the utility of human interventions as local preference supervision for guiding frozen generative
robot policies.
Read the full abstract ▾
All authors Yiqi Tang, Diyuan Shi, Runze Li, Donglin Wang
Materials Science
Orientation engineering is a universal strategy for ferroelectric trans-switching
Cameron Scott and Jorge Íñiguez-González
About two decades ago, “strain engineering” emerged as a powerful strategy to tune epitaxial thin films. Here we contend that ”orientation engineering” - that is, controlling the film's growth direction - constitutes a complementary design route that opens unprecedented opportunities and deserves greater
attention. Specifically, we show that orientation engineering can endow standard ferroelectric compounds with exotic and useful behaviors that may be difficult or impossible to obtain otherwise. Critically, we predict it readily enables ”trans-switching,” where polarization components switch perpendicularly to the applied electric field, paving the way for novel 3D device architectures. We also show that, by controlling the electric screening provided by electrodes in contact with the film, ferrielectric behavior can be obtained. Notably, one can even design an operational mode where the switching is, in effect, fully transversal to the applied field. Supported by phenomenological calculations for representative compounds LiNbO$_3$ and BaTiO$_3$, we thus propose orientation engineering as a powerful and general method to unlock unprecedented functionalities in common ferroelectrics.
Read the full abstract ▾
Robotics
VidAct: Learning Manipulation from In-the-Wild Videos with Object-Centric 3D Awareness
Hang Li, Mingxin Zhang, Zihan Wu et al. (+9 more)
Video demonstrations offer a scalable alternative to costly
robot data for learning manipulation, yet existing reconstruction-based approaches often rely on constrained camera viewpoints or human-to-
robot retargeting, while the reconstructed trajectories are difficult to adapt to new objects configurations without distorting the trajectory shape. Another key limitation is that the resulting policies often lack precise object-level 3D geometry awareness, limiting object grounding and object shape awareness critical for precise manipulation. To bridge these gaps, we propose VidAct, an efficient
video-to-
robot framework that learns object-centric, 3D-aware manipulation policies from a single
monocular video per task and enables zero-shot real-world deployment. VidAct consists of three key components. First, VidAct reconstructs object meshes and motion from arbitrary demo
videos and canonicalizes the motion in the static object frame, avoiding embodiment-specific retargeting and accommodating diverse camera viewpoints. Second, VidAct employ residual trajectory transfer for adapting the reconstructed motion to novel object configurations while preserving its motion shape. Finally, as the key policy-learning component, VidAct predicts simulation-provided privileged complete-object point clouds at each frame as an auxiliary task while retaining RGB-only deployment, providing dense object-centric supervision over both object pose and 3D geometry. Experiments on human,
robot, generated, and internet
videos demonstrate broad
video applicability and zero-shot deployment. Per-frame complete-object 3D supervision improves policy generalization and sim-to-real success, while residual trajectory transfer enables reliable trajectory adaptation with better shape preservation.
Read the full abstract ▾
All authors Hang Li, Mingxin Zhang, Zihan Wu, Yang Tian, Dong Chen, Fengyi Shen, Yuan Meng, Xiangtong Yao, Heng Zhang, Ziyuan Liu, Zhenshan Bing, Alois Knoll
Optimization and Control
Approximation in norm and energy with bounded functions for autonomous variational problems
Pierre Bousquet, Carlo Mariconda and Giulia Treu
We study the approximation of finite-energy Sobolev functions by bounded Sobolev functions for integral functionals associated with possibly signed Carathéodory integrands. The approximating sequence is required to converge strongly in the Sobolev space, while the corresponding energy densities converge in \(L^1\). Ordinary truncations may fail because they replace the gradient by zero on the tails, where the integrand need not be integrable. Our main contribution is a general method based on spatially dependent barriers whose gradients can be adapted to the structure of the Lagrangian.
Read the full abstract ▾
Strongly Correlated Electrons
Quantized Hyperfine Field at an Implanted $\mu^+$ Site in PrPb$_3$: Interplay between Localized $f$ Electrons and an Interstitial Charged Particle
T. U. Ito, W. Higemoto, K. Ohishi et al. (+6 more)
The local effect of an interstitial hydrogen-like particle on localized $f$ electrons was studied in PrPb$_3$ by means of $\mu^+$
spin rotation and relaxation. Spontaneous $\mu^+$
spin precession with harmonic frequencies was observed for the first time in $f$ electron compounds. We demonstrate that the signal is derived from a coupling between the $\mu^+$
spin and the hyperfine-enhanced
nuclear spin of nearest neighbor (nn) $^{141}$Pr with Ising-like anisotropy. The signal also suggests a marked suppression of
spin dynamics of the nn $^{141}$Pr in comparison with that of the bulk $^{141}$Pr. These facts strongly indicate modification of the $f$ electronic state due to the interstitial charged particle.
Read the full abstract ▾
All authors T. U. Ito, W. Higemoto, K. Ohishi, N. Nishida, R. H. Heffner, Y. Aoki, A. Amato, T. Onimaru, H. S. Suzuki
Artificial Intelligence
State Trace Rationale As Auxiliary Task in Reinforcement Learning
Muhammad U. Nasir, Alex Vogt, Steven D. James et al. (+1 more)
We propose STRAT, an auxiliary task that trains deep reinforcement learning (RL)
agents to predict a short textual trace of their own state. Inspired by human spatial
navigation, the description combines landmark, route, and survey knowledge, tracking the
agent's position, inventory, goals, and immediate progress. Environment rules generate this text online without human labelling. Our method adds a single auxiliary head to a standard policy. Across 60 sparse-reward XLand-MiniGrid tasks, STRAT solves complex environments where standard RL fails outright, while compacting state representations and preventing rank collapse. Beyond performance gains, the predicted trace provides a readable account of
agent beliefs at every step for no extra cost.
Read the full abstract ▾
All authors Muhammad U. Nasir, Alex Vogt, Steven D. James, Julian Togelius
Computer Vision
S2T-Unet: A Structure-to-Style Framework for Inter-Modality MRI Translation
Yichao Liu
Inter-modality MRI translation aims to synthesize missing MRI modalities from available acquisitions, reducing the need for additional scanning while preserving clinically relevant anatomical information. However, existing image translation methods often learn intensity mappings without explicitly separating modality-invariant structural information from modality-specific appearance, which may lead to structural information loss or unrealistic image details. In this work, we propose S2T-Unet, a structure-to-style framework that explicitly models these two aspects. Specifically, vector quantization is introduced at the lower-level bottleneck to encode modality-invariant structural information using a learned discrete codebook. At higher levels, a modality transformation module uses decoder features to condition and transform encoder representations toward the target modality, thereby recovering modality-specific intensity and contrast information. Experiments on the IXI multi-contrast MRI
dataset across four translation tasks demonstrate that S2T-Unet is comparable or outperform with state-of-art method.
Read the full abstract ▾
Optimization and Control
Beyond Constant Truncations: The Bounded Sobolev Approximation Gap
Pierre Bousquet, Carlo Mariconda and Giulia Treu
We study the approximation of finite-energy Sobolev functions by bounded Sobolev functions for integral functionals associated with possibly signed Carathéodory integrands. The approximating sequence is required to converge strongly in the Sobolev space, while the corresponding energy densities converge in \(L^1\). Ordinary truncations may fail because they replace the gradient by zero on the tails, where the integrand need not be integrable. Our main contribution is a general method based on spatially dependent barriers whose gradients can be adapted to the structure of the Lagrangian.
Read the full abstract ▾
Machine Learning
Where the Model Changes Its Mind: Hindsight-Divergence Localization for Efficient Reinforcement Learning with Verifiable Rewards
Fanchao Chen, Hengyu Fu, Shivaram Venkataraman et al. (+1 more)
Group-relative methods for reinforcement learning with verifiable rewards (RLVR) learn from differences in rollout outcomes. Independently sampling complete trajectories is costly and does not explicitly explore the decision space at critical positions. Feedback on a completed trajectory can reveal which earlier choices the policy reconsiders, suggesting where to sample alternative continuations. We introduce Hindsight-Divergence Localization (HDL), which uses hindsight-induced changes in
token log-likelihoods to select branch points. HDL generates a small number of complete root trajectories and fills each training group with continuations from the selected positions under the original task context. Each continuation reuses its root prefix and contributes policy updates only through its newly generated suffix, reducing generation cost while focusing additional exploration and learning on decisions after branching. Experiments with three models across math, code, and
agent tasks show gains in both rollout efficiency and task performance. Compared with GRPO at matched group sizes and training steps, HDL yields up to a 2.5$\times$ reduction in generated
tokens and a 1.8$\times$ speedup in rollout wall-clock time. Despite this reduced generation budget, HDL improves performance across all three domains, with gains of up to 12.5 points on
agent tasks.
Read the full abstract ▾
All authors Fanchao Chen, Hengyu Fu, Shivaram Venkataraman, Jiantao Jiao
Cosmology and Nongalactic Astrophysics
Cross-correlating Squared kSZ and HI Intensity Fields: A Map-level ACT–MeerKAT Forecast
Yu-Er Jiang, Yun Liu, Yan Gong et al. (+5 more)
The kinetic Sunyaev–Zel'dovich (kSZ) effect probes the line-of-sight momentum of free electrons, but its blackbody spectrum and velocity-dependent sign make it difficult to isolate and cause conventional cross-correlations with density tracers to vanish. In this work, we investigate the cross-correlation of the squared kSZ field with the projected squared HI intensity field using mock ACT and MeerKAT observations constructed from a Jiutian-1G simulation lightcone. We generate
halo-based kSZ maps and MeerKAT-like HI brightness-temperature cubes over a $13.5^\circ\times13.5^\circ$ field. The simulations incorporate ACT beam smoothing and effective internal-linear-combination noise, together with the frequency-dependent MeerKAT beam, scanning strategy, map-making, and inhomogeneous thermal noise. We analyze signal-only, idealized amplitude-reconstructed, and Wiener-filtered kSZ maps. The resulting squared-field cross-correlations have diagnostic cumulative signal-to-noise ratios of approximately $3$-$4$. Our map-level results demonstrate the detectability of the ${\rm kSZ}^2\times {\rm HI}^2$ correlation in future joint ACT–MeerKAT observations.
Read the full abstract ▾
All authors Yu-Er Jiang, Yun Liu, Yan Gong, Zi-Yan Yuwen, Yin-Zhe Ma, Furen Deng, Xuelei Chen, Qi Guo
Cryptography and Security
Safer Content or Firmer Refusals? A Hybrid Perturbation Defense for Alignment under Harmful Fine-tuning
Muhammad Zeeshan Akram, Mufid Kamel Marican, Anvesh Reddy Yenugu et al. (+2 more)
Fine-tuning-as-a-service lets users adapt a safety-aligned
language model to their own data, but it also creates a harmful fine-tuning attack surface: a small amount of harmful data mixed into an otherwise benign fine-tuning set can degrade the model's
alignment. Two recent
alignment-stage defenses address this problem at different levels of the model. Vaccine improves the robustness of hidden
embeddings to the representation shifts induced by harmful fine-tuning, whereas Booster simulates harmful weight updates and attenuates their effect during
alignment. We investigate whether these mechanisms are complementary and propose VaccineBooster, a single
alignment procedure that combines
embedding perturbation and weight-level gradient attenuation within each training step. On Llama-2-7B aligned with BeaverTails and then attacked through poisoned fine-tuning, VaccineBooster achieves the lowest OpenAI moderation score among the compared defenses, 0.315, while a Booster-Only variant retains the highest post-attack refusal rate, 50%. Together with ablations over the
embedding-perturbation and gradient-attenuation strengths, these results indicate a trade-off:
embedding perturbation primarily reduces flagged harmful content, whereas gradient attenuation primarily preserves explicit refusal behavior. Because our evaluation uses ten
prompts and a single unseeded run per configuration, we report this trade-off as an observed pattern rather than a statistically resolved effect. These results provide practical guidance for prioritizing content safety or refusal retention when aligned models are exposed to untrusted fine-tuning.
Read the full abstract ▾
All authors Muhammad Zeeshan Akram, Mufid Kamel Marican, Anvesh Reddy Yenugu, Ali Zain Kaimkhani, Minghong Fang
Quantum Physics
Exact Trotterization in the Degenerate Quantum Rabi Model
I. Lizuain and A. Rodriguez-Prieto
Digital
quantum simulation of the
quantum Rabi model in the deep strong coupling regime motivates the need for an exact characterization of Trotterization errors, so far known only through qualitative bounds. Here we derive such a characterization in the exactly solvable degenerate limit, valid for arbitrary coupling strength. The first-order Trotter operator admits a closed expression, showing that the generated phase-space trajectory lies exactly on a circle that is rescaled and rigidly rotated with respect to the exact one, rather than merely approximating it. This
geometric renormalization yields a closed form for the state fidelity, structurally identical to the physical survival probability governing revivals in the deep strong coupling regime: both are the same
geometric question – the distinguishability of a displaced Fock state – applied to two different pairs of points in phase space. We further show that this
geometric error is exactly correctable by promoting the Trotter-step coupling to a complex amplitude, at no additional
circuit cost. These results, obtained in the exactly solvable degenerate limit, provide a
benchmark for Trotterization errors in the general Rabi model.
Read the full abstract ▾
Artificial Intelligence
IronLLM: Forging Compact Edge-Native Language Models for Real-Time Embodied Intelligence
Changdi Yang, Fengquan Jiao, Haochih Lin et al. (+14 more)
We present IronLLM-0.6B, a 654M-parameter
language model designed for efficient on-device inference. IronLLM-0.6B combines a hybrid
attention architecture with X-MTP, a lightweight shared-KV multi-
token prediction design that eliminates per-depth KV-cache replay and employs a lightweight verification head for rollback-free drafting, achieving a 1.48x decoding speedup. The model is pretrained on approximately 6.2 trillion
tokens using a quality-oriented data pipeline and is further post-trained with Multi-Domain On-Policy
Distillation to integrate capabilities from domain-specialized teachers. To better meet the low-latency requirements of on-device scenarios, IronLLM-0.6B adopts an Instruct-Only design. Evaluations show that IronLLM-0.6B achieves competitive performance relative to larger models such as Qwen3.5-0.8B and MiniCPM5-1B, while producing more concise responses on many tasks. We further present IronLLM-0.6B-Light, which replaces RMSNorm with Dynamic Tanh and simplifies several computationally expensive components to improve inference and quantization efficiency. Together, the IronLLM models provide an effective performance-efficiency trade-off for resource-constrained deployment.
Read the full abstract ▾
All authors Changdi Yang, Fengquan Jiao, Haochih Lin, Haoran Yang, Jing Xiao, Liangyu Huo, Suxin Lu, Tiance Chen, Wei Liu, Yinggan Xu, Yunxiang Lu, Zai Zheng, Zhirui Xie, Zhongyang Che, Ziyan Tang, Zuoxiang Zhao, Jian Yao
Machine Learning
Markovian Nonconvex ADMM for Reinforcement Learning: Bellman-Resolvent Stability Beyond Smooth Blocks
Zhaojun Peng
We identify and study a structural mechanism for Markovian nonconvex ADMM in reinforcement learning. Using finite discounted MDPs as a canonical proving ground, we show that the discounted Bellman resolvent $(I-\gamma P_\pi)^{-1}$ can provide the multiplier stability that classical nonconvex ADMM analyses often obtain from a designated smooth block. Starting from this mechanism, we establish convergence under controlled Markov sampling and then under
stochastic observations using an empirical Bellman surrogate that jointly represents the random residual and its Jacobian. Markov mixing, initialization drift, observation noise, and decaying bias enter as one operator perturbation, avoiding unbiased product and double sampling requirements. When the perturbations are square summable, the true KKT residual converges almost surely to zero. Under a finite conditional fourth moment condition, a companion iterate satisfies $ \mathbb{E}[\widetilde G_{K+1}] \le A/T+(B/T)\sum_{k<T}m_k^{-1}, $ which becomes $O(T^{-1}+T/N)$ for total Markov sample budget $N$, giving $O(\epsilon^{-1})$ iteration complexity and $O(\epsilon^{-2})$ sample complexity for squared KKT accuracy $\epsilon$. Beyond stationarity, discounted occupancy coverage yields $J^\star-J(\pi)=O(\sqrt G)$ for direct tabular policies, so covered exact KKT points are globally optimal, while a statewise quadratic Bellman-improvement condition sharpens the relation to $O(G)$. Finally, nonlinear policy, projected Bellman, and explicit occupancy formulations exhibit the same chain of operator invertibility, dual representation, and multiplier stability. This supports discounted operator invertibility as a reusable structural principle for primal-dual reinforcement learning.
Read the full abstract ▾
Systems and Control
Receding Horizon Control and Dissipativity - Optimal Control, Games and Uncertainty
Sophie Hall, Jonas Schießl, Sergio Grammatico et al. (+1 more)
We provide a tutorial overview of receding horizon control across deterministic,
stochastic, multi-
agent, and game-theoretic settings. We contrast the minimizer of an optimal control problem with the
equilibrium of a game in terms of cost and constraint handling to motivate the difference of MPC schemes involving multiple
agents/players. Interestingly, dissipativity theory and the turnpike property have turned out to be a unifying thread and system-theoretic backbone across all these schemes. This paper is the first to give an overview of results, summarizing fundamental insights, drawing parallels, and highlighting technical differences in dissipativity and turnpike analysis across economic, game theoretic, and
stochastic settings. We also explore recent trends in numerical methods for games. We conclude by pointing to a list of open challenges.
Read the full abstract ▾
All authors Sophie Hall, Jonas Schießl, Sergio Grammatico, Timm Faulwasser
Instrumentation and Methods for Astrophysics
LUMA: A CNN for Strong Gravitational Lens Searches in Astronomical Imaging
Giovanni Vincenzo Donatiello, Achille A. Nucita, Francesco De Paolis et al. (+2 more)
We present LUMA, a convolutional neural network (CNN) pipeline for the automated detection of strong gravitational lenses in simulated astronomical imaging. The method combines a physically motivated preprocessing stage, which enhances faint arc and ring features, with a compact three-block CNN trained using class reweighting and modern learning-rate scheduling. On simulated data, the model reaches test accuracies of about $96$% and receiver operating characteristic (ROC) area-under-the-curve (AUC) values of $\simeq 0.99$ for the non-trivial classes, while confusion-
matrix analysis shows high completeness and purity for lens candidates. These results demonstrate that relatively lightweight CNN architectures can provide a competitive baseline for strong-lens searches, and they motivate future extensions toward real survey images and
transformer-based models.
Read the full abstract ▾
All authors Giovanni Vincenzo Donatiello, Achille A. Nucita, Francesco De Paolis, Antonio Franco, Francesco Strafella