Computer Vision
LVS: Local View Synthesis from Relative Camera Pose by Reusing Previous Views
Qizhou Huo, Xuan Sun, Yongfei Guo et al. (+2 more)
Interactive scene exploration requires frequent view updates, although small camera motions preserve much of the visible content. Conventional 3D Gaussian Splatting nevertheless renders each target view, leaving this image overlap unexploited. Reusing rendered images offers an alternative.
Geometric warping alone cannot recover newly exposed content and remains sensitive to depth errors. We propose a per-scene framework that replaces repeated scene rendering for nearby views with relative-pose-guided RGB-D image reuse.
Geometric warping uses depth and relative pose to transport source content, while a lightweight multiscale network predicts RGB residuals to correct artifacts and infer missing appearance. Cached source features further reduce repeated computation. On GS-render, residual refinement improves PSNR by 0.72~dB over pure warping; evaluations on captured and rendered scenes demonstrate low query latency. This separation of scene rendering from local view updates supports responsive scene exploration, with potential applications in augmented and virtual reality.
Read the full abstract ▾
All authors Qizhou Huo, Xuan Sun, Yongfei Guo, Zhipeng Wang, Yuanhao Gong
Robotics
SuperNav: An Agentic Navigation System for Any Task in Any Scene
Jinkai Zhang, Jingyi Xu, Yuanhong Yu et al. (+5 more)
General-purpose service
robots need
navigation systems that can handle diverse human requests in unfamiliar environments, combining task generality with scene generality. Some existing methods fine-tune
multimodal large
language models (MLLMs) to predict
navigation actions, making their behavior dependent on the coverage of
navigation training data and potentially limiting generalization to new requests and environments. Our key insight is to let the MLLM focus on interpreting requests, understanding scenes, and making decisions while preserving its general-purpose capabilities and delegating motion execution to
navigation tools. To realize this idea, we introduce SuperNav, which equips a pretrained MLLM with a specialized
agent harness without
navigation-specific fine-tuning of the MLLM. Our harness supports these decisions with
Navigation Skills,
agent-oriented Tools for physical interaction, and task-progress and context management. A unified visual-point interface connects decision-making to motion by allowing the model to specify destinations directly in images and revise its decisions from execution feedback. Together, these components support sustained
navigation across different task requirements and environments. SuperNav outperforms four evaluated baselines on instance-level, multi-object, and demand-driven tasks. Category-level evaluation on HM3D and deployment on a real quadruped
robot further demonstrate its applicability across environments. Project Page: https://zju3dv.github.io/SuperNav/
Read the full abstract ▾
All authors Jinkai Zhang, Jingyi Xu, Yuanhong Yu, Jiarui Guo, Ruizhen Hu, Hujun Bao, Xiaowei Zhou, Sida Peng
Computers and Society
The learner who does not learn: when optimizing a pedagogical metric degrades LLM tutoring
Daniel Domínguez Figaredo and Rafael Fernández De la Cruz
It is assumed that a natural way to improve the pedagogical quality of large
language model tutors is to define a metric of instructional performance and fine-tune the model against it. To test this strategy, we designed a metric of pedagogical adaptivity that scores each instructional decision in a learning sequence against the conditions of the learning situation, which is the standard used for automated pedagogical scoring. We audited a frontier tutor across 2,000 learner scenarios, corrected its weakest cases by fine-tuning an open-weights proxy, and asked 31 trained educators to rate the pedagogical
alignment of the outputs blind, before and after correction. The metric increased from +0.05 to +0.42 for the corrected cases, while the expert ratings decreased from 4.46 to 3.03, with the unmodified controls remaining unchanged and a base-proxy control ruling out the change of model. The tutor performed worse because any metric that scores decisions independently and averages them is maximized by repeating the single best decision, and the fine-tuned model collapsed to that exact optimum in every case, in and out of sample. Educators identified the repetition, which such metrics cannot represent, and preserving the learner's trajectory in the score reduced, but did not reverse, the metric's verdict. Weight analysis traced the correction to the model's output projection, where it had memorized its training strings rather than learned to adapt. We conclude that measurement validity does not imply optimization validity, and we derive design principles for
benchmarks that assess or train AI-tutor instruction.
Read the full abstract ▾
Artificial Intelligence
Use and Disuse: Intent-Structured Experience Consolidation for Memory and Learning in LLM Agents
Xiangyi Zeng, Baihang Liu, Xutong Wang et al. (+3 more)
The
evolution of Large
Language Model
agents from single-task execution to long-term autonomous operation highlights the critical challenge of transforming continuous experiences into reusable knowledge. To address this, we propose Hippocam, a hierarchical
memory and continual learning architecture. Hippocam draws inspiration from two characteristics of human
memory: cognitive processes selectively maintain information relevant to current goals, while long-term
memories form gradually through repeated consolidation. Accordingly, Hippocam structures an
agent's ongoing work as nested intents. The active context remains centered on the current intent, while completed intents are consolidated into the task-relevant outcomes and state needed for subsequent work, rather than carrying forward their full working details. Concurrently, a recursive prefix consolidation mechanism repeatedly consolidates earlier history, causing long-unused experiences to become increasingly abstract. Original interactions are preserved, allowing the
agent to progressively recover finer-grained details through the hierarchy and stop once sufficient information is available. Crucially, when past experiences are recalled and reintegrated into active work, they undergo subsequent consolidation alongside new experiences, thereby being reinforced, supplemented, and updated. Through this
memory dynamic of use and disuse, Hippocam connects working context, long-term
memory, knowledge accumulation, and skill learning within a single continuously evolving experiential process. This enables
agents to learn and evolve capabilities through their own experiences without parameter updates.
Read the full abstract ▾
All authors Xiangyi Zeng, Baihang Liu, Xutong Wang, Ze Jin, Yunpeng Li, Qixu Liu
Robotics
PIER: An Evidence-Gated Execution Interface for Robotic Manipulation
Zoe Li, Anze Wang, Zhongyu Chen et al. (+1 more)
Generating a plausible
robot action does not establish that current observations justify its execution. Motivated by exploratory observations of high-confidence visual outputs under severe occlusion, we present PIER, an execution-authorization interface that separates evidence checks, decision provenance, and stage-scoped re-observation from hardware control. The deterministic gate evaluates declared visual and tactile inputs, while its caller maintains a budget of at most one re-observation per stage. We evaluate the implementation using 1,600 threshold-grid cases and 1,200 paired synthetic traces spanning score noise, missing tactile inputs, stale observations, and falsely reassuring scores. The finite grid yields zero declared invariant violations, and a matched Boolean baseline reproduces all non-recovery decisions. Under synthetic score noise, re-observation reduces valid-state denials from 57/120 to 18/120 while increasing invalid-state proceeds from 5/120 to 7/120. Stale and falsely reassuring inputs expose limitations that threshold checks alone cannot resolve. Exploratory visual, tactile, and
robot setup records provide context but do not establish physical task performance. These results characterize an inspectable authorization interface and its input-contract limitations, without claiming superiority over equivalent rule logic, calibrated tactile accuracy, or certified physical safety.
Read the full abstract ▾
All authors Zoe Li, Anze Wang, Zhongyu Chen, Jingran Hu
Combinatorics
Lower bounds for Ramsey numbers: $\mathrm{R}(6,8)\ge 135$ and $\mathrm{R}(8,10)\ge 345$
Fritz Cremer
We prove the lower bounds $\mathrm{R}(6,8)\ge 135$ and $\mathrm{R}(8,10)\ge 345$, improving the bounds 134 and 343 listed in the April 2026 revision of Radziszowski's dynamic survey. We give explicit red/blue colorings of $K_{134}$ with no red $K_6$ or blue $K_8$, and of $K_{344}$ with no red $K_8$ or blue $K_{10}$, and verify them with two independently written exhaustive clique checkers. Starting from published colorings, we find these witnesses by adding vertices and repairing the resulting monochromatic cliques through local search. The search uses exact conflict counts, preparation moves aimed at making remaining cliques cheaper to break, and mutations followed by repair. A loss based on the largest monochromatic clique containing each edge yielded a useful intermediate state for the $\mathrm{R}(6,8)$ construction. Guided by the author, an AI coding
agent wrote and ran the search code; we describe the methods and document the ancestry of the resulting colorings.
Read the full abstract ▾
Astrophysics of Galaxies
Too many or too bright? Status and perspectives in the study of the UV Luminosity Function at z>10
A. Fontana, H. Atek, V. Bromm et al. (+20 more)
The finding that
galaxies in the first 500Myr were much more numerous and/or brighter than expected is often cited as one of the main discoveries of the James Webb Space
Telescope in its first four years of operation. This result is now solidly based on an impressive number of surveys that have detected, spectroscopically confirmed, and characterized a relatively large sample of
galaxies up to z=14.44, the
redshift of the most distant
galaxy currently known. This finding has motivated the introduction of additional physical mechanisms into theoretical models to reproduce the observed numbers and luminosities. These results contrast with the failure, despite ongoing efforts, to confirm any
galaxy candidate beyond this
redshift. This review summarizes the status of the search for and characterization of
galaxies at z>9, focusing on the
evolution of the UV Luminosity Function (UVLF), which is the main statistical description of the
galaxy population that can be obtained at these
redshifts. We first provide a comprehensive overview of the full set of JWST data obtained so far. We perform a 'meta–analysis' on the entire data set and provide a simple yet effective description of the global
evolution of the UVLF, averaging over the various surveys. We also compare these results extensively with theoretical models that were developed to match the new JWST observations and that explore different physical mechanisms able to drive the observed
evolution. We conclude with a summary of the contrasting results regarding the search for
galaxy candidates at z>15, and outline the requirements for future surveys aimed at breaking this barrier.
Read the full abstract ▾
All authors A. Fontana, H. Atek, V. Bromm, A. J. Bunker, M. Castellano, P. Dayal, F. D'Eugenio, D. Elbaz, A. Ferrara, S. L. Finkelstein, S. Fujimoto, M. Giavalisco, Y. Harikane, C. Kobayashi, F. Pacucci, P. G. Pérez-González, P. Rosati, A. Trinca, R. Tripodi, P. Santini, E. Vanzella, A. Venditti, A. Weibel
Commutative Algebra
Derived functors and Hilbert polynomials over dominant Cohen-Macaulay rings
Tony J. Puthenpurakal
Let $(A,\mathfrak{m})$ be a Cohen-Macaulay local ring of dimension $d$, residue field $k$ and let $I$ be an $\mathfrak{m}$-primary ideal. Assume $k$ is infinite. Let $N$ be a perfect $A$-module of dimension $t \geq 1$. Let $M$ be a MCM $A$-module. The function $n \rightarrow \ell(\text{Tor}^A_j(M, N/I^{n+1}N))$ is of
polynomial type and let $r_{I, N}^j(M)$ be its degree. In general we have $r_{I, N}^j(M) \leq r_{I,N}^2(\text{Syz}^d_A(k))$ for $j \geq d + 2$. If $A$ is also dominant in the sense of Takahashi and $M$ is non-free then we show $\limsup_{j \rightarrow \infty} r_{I, N}^j(M) = r_{I,N}^2(\text{Syz}^d_A(k))$. We prove an analogous result when $V$ is a Cohen-Macaulay module of finite injective dimension and the
polynomial type function $n \rightarrow \ell(\text{Ext}_A^j(M, V/I^{n+1}V))$
Read the full abstract ▾
Machine Learning
Using Weisfeiler-Leman Features for Algorithm Selection in Constraint Optimisation
Alessio Pellegrino and Jacopo Mauro
Algorithm Selection is essential for efficient Constraint Programming. Over the years, many
algorithm selectors based on machine learning methods have been successfully applied, yet traditional feature extraction methods often rely on manually decided instance-level statistics that fail to capture the underlying problem structure. In this paper we aim to bridge this gap by introducing a novel, automated feature extraction methodology that integrates graph conversion and Weisfeiler-Lehman graph kernels to generate robust structural representations of problem instances. The 1-WL test bounds the graph-distinguishing power of standard message-passing Graph Neural Networks (GNNs), and suitable GNN architectures match this bound \citep{Xuetal2018}. WL-based features offer an alternative that does not require training a GNN. Our primary contribution is a cut-based representation (WLc) designed to model structural partitions and provide a more nuanced predictive signal. We evaluate our approach on instances from the 2023–2025 MiniZinc Challenges across two tasks: maximizing Borda count scores and maximizing predictive accuracy. Experimental results across Support Vector Machines, Random Forests, and Multi-Layer Perceptrons demonstrate that cut-based features outperform fzn2feat with SVMs, while results with RFs and MLPs are closer.
Read the full abstract ▾
Computation and Language
A persistent accuracy ceiling in automated verbal deception detection
Riccardo Loconte, Jonas Festor, Zane Fatjanova et al. (+2 more)
Automated methods have been proposed to overcome the limitations of human verbal deception detection, but evidence remains fragmented across disciplines. We systematically reviewed 25 years of research (289 reports, 6,136 classification models) and meta-analyzed 3,653 models nested within 97
datasets. Pooled accuracy was 74.4% (95% CI: 71.2%-77.4%) with substantial heterogeneity. Accuracy was driven by methodological quality (ground truth, data source, class balance, evaluation procedure) more than by model complexity: the adoption of
embeddings and large
language models has not translated into improved predictive performance. Only 12.46% of reports used data with verifiable ground-truth, and only 23.96% of models were evaluated on independent data. The pooled accuracy aligns with meta-analyses of manual approaches, suggesting a ceiling of 70-75%, unlikely to be lifted by current research conventions.
Read the full abstract ▾
All authors Riccardo Loconte, Jonas Festor, Zane Fatjanova, Mariam Bolkvadze, Bennett Kleinberg
Solar and Stellar Astrophysics
EWOCS-XI: Anisotropic expansion and shear in the cluster Westerlund 1. A 2D kinematic analysis with Gaia DR3
L. Prisinzano, F. Damiani, C. Ordenes-Huanca et al. (+14 more)
Westerlund 1 (Wd1) is one of the Milky Way's most massive young clusters, in which heavy extinction and field contamination have historically hampered its kinematic characterization. We aim to assess Wd1's internal kinematics, dynamical state, and formation scenario. Using a robust 5D clustering approach, we identified 559 bona fide members. We modeled their proper motions via a 2D first-order Taylor expansion to derive the velocity gradient tensor, mapping expansion, rotation, and shear. Restricting the kinematic analysis to 524
stars with proper motion uncertainties $<0.5~\mathrm{mas\,yr^{-1}}$, we find at least 103 gravitationally unbound members, including 14 walkaway
stars. The 2D analysis reveals a statistically significant divergence (>4$\sigma$) highlighting preferred expansion along the
Galactic latitude (position angle $164^\circ\pm10.1^\circ$). A detected cross-gradient ($\partial v_b / \partial l = -0.41 \pm 0.08$ km s$^{-1}$ pc$^{-1}$) provides a clear signature of shear on the plane of the sky. Wd1 does not behave as a single cohesive population, but comprises two decoupled subsystems: a stable, bound core in dynamical
equilibrium and an escaping, unbound envelope. This prominent anisotropic expansion perpendicular to the
Galactic plane has a kinematic timescale ($\Delta t_{\mathrm{kin}} = 2.08 \pm 0.44$ Myr) matching the independent orbital mid-plane crossing time ($\Delta t_{\mathrm{orbit}} = 2.10 \pm 0.05$ Myr). This temporal
alignment indicates the expansion onset coincided with Wd1's
Galactic disk passage, triggered either by internal dynamics (e.g., early
supernova feedback) or external
tidal perturbations. Finally, the robust unbound population demonstrates that active cluster dissolution is underway, driven by discrete dynamical ejection events rather than long-term
stellar evaporation.
Read the full abstract ▾
All authors L. Prisinzano, F. Damiani, C. Ordenes-Huanca, M. Tarantino, S. Sciortino, N.J. Wright, V. Almendros-Abad, M. Andersen, A. Bayo, R. Bonito, V. Cusimano, M. Haberle, K. Mužic, I. Negueruela, E. Sabbi, A. Winter, M.G. Guarcello
General Topology
The size of the poset of compatible locally quasi-convex topologies on locally compact abelian groups
ZhouXiang Huang
For a Hausdorff locally quasi-convex abelian group $G$, let $\C(G)$ be the poset of all Hausdorff locally quasi-convex group
topologies on its underlying group having the same continuous characters as $G$. We prove that, whenever $G$ is non-precompact, $\C(G)$ contains an order-isomorphic copy of $(\Pow(\cont),\subseteq)$, where $\cont=2^{\aleph_0}$. The
embedding takes values between the Bohr
topology and the original
topology. Consequently, every infinite discrete abelian group $D$ satisfies $|\C(D)|=\width\C(D)=2^{2^{|D|}}$. The discrete reduction for locally compact abelian groups then yields exact cardinality and width formulas for all such groups; in particular, both invariants equal $2^{\cont}$ for every non-compact $\sigma$-compact locally compact abelian group. These results answer Questions 6.1–6.3 and 6.5–6.7, and the locally compact case of Question 6.4, posed by L.~Außenhofer and D.~Dikranjan in \cite{AD20}. The
embedding also applies to non-compact complete metrizable locally quasi-convex groups. A non-compact precompact
nuclear group with a unique compatible
topology shows that non-compactness alone does not suffice in the
nuclear setting. Finally, the compatible poset of $\R^{\N}$ is not order-isomorphic to that of any discrete abelian group.
Read the full abstract ▾
Machine Learning
Credal Machine Learning for Risk-Averse Decision Making
Timo Löhr, Paul Hofman, Maximilian Muschalik et al. (+1 more)
In many machine learning applications, it is necessary to guard against worst-case scenarios and predictions that could result in substantial losses. In principle, this can be achieved by training risk-averse predictive models that minimize loss functions such as conditional value-at-risk (CVaR), rather than relying on models that perform well on average. In practice, however, the effectiveness of this approach to risk aversion is undermined by the learner's uncertainty regarding the true loss distribution and, consequently, the true CVaR. To achieve reliable risk-aversion, we propose a method in which this (epistemic) uncertainty is represented in terms of credal sets, i.e., sets of probability distributions. More specifically, we develop an efficient yet reliable learner that produces predictions in the form of credal sets and combine it with a novel decision rule that maps each credal set to a single predictive distribution for CVaR minimization. Across classification, under distribution shift, and in reinforcement learning, our approach reliably avoids catastrophic decisions, while sacrificing little in expected performance.
Read the full abstract ▾
All authors Timo Löhr, Paul Hofman, Maximilian Muschalik, Eyke Hüllermeier
Artificial Intelligence
Universal Textual Teaching for LLMs
Zhanyi Lu and Huan Wang
Knowledge
distillation (KD) transfers knowledge from stronger Teacher models to weaker Student models, but most methods require training the Student parameters, thereby binding the distilled knowledge to a specific architecture and checkpoint. This implicit representation is difficult to interpret or reuse across models and limits KD for API-only or costly-to-train models. This paper studies knowledge transfer for large
language models (LLMs). We introduce Universal Textual Teaching (UTT), a parameter-update-free framework that distills observed Teacher-Student knowledge gaps into a textual, interpretable, and reusable natural-
language artifact called Primer. Specifically, UTT first identifies representative gap cases through paired evaluations, and iteratively updates the Primer via multi-role interactions: the Student attempts each task, the Prompter turns evaluation feedback into a teaching instruction, the Teacher provides a targeted demonstration, and the Synthesizer consolidates validated lessons. Empirically, on the challenging math (Omni-MATH-2) and code generation (KernelBench) tasks, extensive results confirm the effectiveness of the method: UTT remarkably raises the Student's accuracy from 9.4% to 48.6% and Fast1 accuracy from 9% to 35% on KernelBench, while increasing mathematical
reasoning accuracy from 27.6% to 51.7%. UTT also performs better than representative
prompt engineering and parameter-based KD methods. Of note, UTT is shown to be generalizable across different Teachers and Students: a Primer synthesized for one Teacher-Student pair can generalize to other Students that do not participate in the synthesis.
Read the full abstract ▾
Machine Learning
A Geometric Approach to Soft Actor-Critic with Zonotopes for Locomotion Learning
Panagiotis Roditis, Panagiotis P. Filntisis and Petros Maragos
Off-policy actor–critic methods control overestimation bias by taking the minimum of two critics. This uses the same aggregation rule everywhere, regardless of how the critics disagree. We propose GeZo-SAC, which uses auxiliary
geometric representations to adapt critic pessimism to the state and action. Alongside its scalar value, each critic predicts a set of generators defining a zonotope. Probing this zonotope along sampled directions provides a
geometric width, "subtracted from each critic value as a pessimistic offset, and a measure of disagreement between the two critics, aggregated with log-sum-exp. This disagreement controls how the critics are combined, moving from a width-weighted average toward the usual minimum as disagreement increases. At inference, the deployed policy is an unmodified SAC actor, since the generators are used only on the critic side during training.Across four MuJoCo-v5 locomotion
benchmarks and six off-policy baselines, GeZo-SAC achieves the highest mean return on Ant-v5 and Hopper-v5 and remains competitive with other methods on the remaining tasks. Our analysis further shows that GeZo-SAC achieves the lowest average actuator work and action effort per metre among the evaluated methods, while maintaining near-zero measured overestimation frequency across all four environments.
Read the full abstract ▾
Quantum Physics
The Imaginary Component of the Weak Momentum: Spectral Vanishing and Gauge Invariance
A. N. Khondker
We establish a rigorous functional-analytic framework for the imaginary component $p_I$ of the complex weak momentum $p_Q = p_R + ip_I = \frac{\hbar}{i}\nabla\ln\psi$. For every wavefunction in the Sobolev space $H^1(\Omega)$, including those with nodal sets, we prove that the expectation value of $p_I$ vanishes identically, as a consequence of the regularity of $\rho = |\psi|^2$ and without appeal to the polar form $\psi = Re^{iS/\hbar}$. We further show that \textnormal{(i)} the Bohm
quantum potential arises entirely from $p_I$ and satisfies $\langle V_{\mathrm{qu}}\rangle = \langle p_I^2\rangle/2m$, from which the known Fisher information relation follows as a corollary; \textnormal{(ii)} the group velocity of a Bloch band is the density-weighted average of $p_R/m$, with no contribution from $p_I$; \textnormal{(iii)} the decomposition is gauge-invariant under
magnetic minimal coupling, the vector potential acting only on $p_R$; and \textnormal{(iv)} the results extend to curved Riemannian
manifolds.
Read the full abstract ▾
Emerging Technologies
RAPID-DNN: Reliability-Aware Partitioning for Robust DNN Inference Deployment on Edge AI
Mukta Debnath, Krishnendu Guha, Debasri Saha et al. (+2 more)
Deep neural networks (DNNs) deployed on heterogeneous edge accelerators are increasingly exposed to runtime hardware faults. Tight power and resource budgets push these platforms toward aggressive voltage scaling and limit the use of hardware fault mitigation, while process variation, thermal stress, and device aging raise fault rates further. The resulting transient bit level corruptions propagate through inference and can sharply reduce prediction accuracy. Existing DNN partitioning frameworks optimize latency and energy under fault free assumptions, so their deployment strategies often fail to sustain reliable inference in realistic operating conditions. This paper presents RAPID DNN, a reliability aware partitioning framework that adds runtime reliability as a third optimization objective alongside latency and energy. RAPID DNN first characterizes layer wise reliability through systematic fault injection in the weight and activation domains to identify vulnerability critical layers. The resulting sensitivity profile guides a three objective NSGA II optimizer that jointly minimizes inference latency, energy consumption, and expected fault induced accuracy degradation. We evaluate RAPID DNN on AlexNet, SqueezeNet, ResNet18, VGG16, and MobileNetV2 using heterogeneous Eyeriss and SIMBA accelerator profiles. Across fault models and injection probabilities, it consistently improves inference robustness over the fault agnostic baseline. Under the representative mixed random fault configuration, RAPID DNN raises average Top 1 accuracy by 9.81 percentage points and reduces average expected fault induced accuracy degradation by 36.86%, at a cost of 2.38% average latency overhead and 3.34% average energy overhead. These results show that treating reliability as a partitioning objective enables robust DNN deployment on heterogeneous edge accelerators under runtime hardware faults.
Read the full abstract ▾
All authors Mukta Debnath, Krishnendu Guha, Debasri Saha, Amlan Chakrabarti, Susmita Sur-Kolay
Optimization and Control
Adaptive dynamic programming using Lyapunov function constraints
Thomas Göhrt, Pavel Osinenko and Stefan Streif
This work is concerned with a stabilizing adaptive dynamic programming (ADP) approach to approximate solution of a given infinite-horizon optimal control problem. Since the latter problem cannot, in general, be solved exactly, a parametrized function approximator for the infinite-horizon cost function is introduced in ADP (so called “critic”). This critic is used to adapt the parameters of the function approximator. The so called “actor” in turn derives the optimal input of the system. It is a notoriously hard problem to guarantee closed-loop stability of ADP due to the use of approximation structures in the control scheme. Since at least stabilizability is always assumed in the analyses of ADP, it is justified to invoke a respective Lyapunov function. The proposed ADP scheme explicitly uses the said Lyapunov function to simultaneously optimize the critic and guarantee closed-loop stability. A Hessian-free optimization routine is utilized for the actor and critic optimization problems. Convergence to prescribed vicinities of the optima is shown. A computational study showed significant performance improvement for the critic-based approach compared a nominal stabilizing controller for a range of initial conditions.
Read the full abstract ▾
Combinatorics
Sharp linear Turán remainders for cycle and edge extremality
Jifu Lin, Xiaolin Wang and Guangmiao Yu
For a graph $G$, let $e(G)$ denote the number of edges of $G$, and let $c(G)$ be the number of distinct cycles in $G$. Morrison, Roberts and Scott asked whether, for every fixed graph $H$ and all large $n$, some $n$-vertex $H$-free graph maximizes both $e(G)$ and $c(G)$. In this paper, we show that the answer is no for every possible chromatic number. Let $T_{n,r}$ be the complete $r$-partite Turán graph on $n$ vertices. We find a constant $\gamma_r>0$ such that if $n$ is sufficiently large and $\mathcal{H}$ is a finite family with $\chi(\mathcal{H})=r+1$ satisfying \[ ex(n,\mathcal{H})<e(T_{n,r})+\gamma_r n, \] then every cycle-maximal $\mathcal{H}$-free graph is edge-extremal, and $\gamma_r$ is best possible.
Read the full abstract ▾
Differential Geometry
Mollifier smoothings of strongly convex $C^0$-Finsler structures
Ryuichi Fukuoka and Anderson Macedo Setti
Let $M$ be a smooth
manifold and $TM$ its tangent bundle. A $C^0$-Finsler structure of $M$ is a continuous function $F: TM \to [0, \infty)$ such that restricted to each tangent space is an asymmetric norm. A mollifier smoothing of $F$ is a family of Finsler structures $F_\varepsilon: TM \to [0, \infty)$, parameterized by $\varepsilon > 0$ and constructed using the standard mollifier, such that $F_\varepsilon \to F$ uniformly on compact subsets as $\varepsilon \to 0$. In this work, we construct two mollifier smoothings of $F$. In the first, we assume that $F: TM \to [0, \infty)$ is strongly convex, that is, that $F$ restricted to each tangent space is strongly convex. In the second, we assume that $M=G$ is a Lie group and that $F:TG \to [0, \infty)$ is strongly convex and left-invariant, and in this case, $F_\varepsilon$ is also left-invariant. In both cases, we prove that if $F$ is a Finsler structure, objects such as the fundamental tensor, the Chern, Cartan, Hashiguchi and Berwald connections, and the flag
curvature of $(M, F_\varepsilon)$ converge uniformly on compact subsets to the corresponding objects of $(M, F)$ as $\varepsilon \to 0$.
Read the full abstract ▾
Computer Vision
ContourVLA: A Closed-Loop Perception-Action Contour Policy for Generalized Referring Expression Segmentation
Ruicheng Zhang, Kaiwen Shen, Jiaqi Hou et al. (+6 more)
Generalized referring expression
segmentation (GRES) requires dynamically balancing high-level semantics for identifying a variable number of
language-specified referents with fine-grained visual evidence for precise boundary delineation. This requirement challenges existing cascaded
vision-
language architectures, which typically rely on static feature interfaces and single-pass mask prediction, limiting adaptive perception and
geometric correction. We introduce ContourVLA, a
vision-
language-action policy that recasts GRES as a closed-loop visuomotor process, in which editable contours serve as explicit policy states that condition
multimodal perception and are updated by
geometric action chunks.
Evolution-Aware Semantic Scheduling (EASS) couples contour-guided bidirectional boundary sampling with state-conditioned routing of multilevel
multimodal features, adapting perception to each contour state. Following supervised initialization, Dustbin-Augmented Entropic Credit Transport GRPO (DECT-GRPO) jointly optimizes discrete grounding and continuous contour actions with instance-level credits. Its rollout rewards and credits are derived from soft prediction-target correspondences that account for false positives and missed targets. ContourVLA improves gIoU over the strongest evaluated baselines by 8.7, 2.8, and 2.7 points on gRefCOCO val, testA, and testB, respectively, and achieves the highest mIoU across all eight RefCOCO, RefCOCO+, and RefCOCOg splits.
Read the full abstract ▾
All authors Ruicheng Zhang, Kaiwen Shen, Jiaqi Hou, Shuhan Yang, Junchao Huang, Kewei Zhang, Jun Zhou, Li Jiang, Shen Zhao
Cryptography and Security
Could LLM Watermark Detection be Public?
Georgios Milis, Tom Sander, Tomáš Souček et al. (+2 more)
Watermarking large
language models is popular for tracing chatbot and agentic outputs, yet detectors remain unreleased since exposing them could let attackers do targeted edits with the detector's feedback. However, watermarks are already vulnerable to uninformed tampering attacks. We thus first quantify whether a public detector would be an additional liability in a deployment setting at varying levels of access, from
token-level scores to a binary verdict. Second, we introduce a split-key public-private watermarking method that exposes one key through a public detector while keeping the other for full verification and forensics. An informed attacker can only move the public signal, creating an imbalance between public and private scores. We introduce a statistical test for this imbalance, and combine it with the full key verdict in a two-stage mechanism. Third, we evaluate the split-key method on a wide range of removal and forgery attacks, comparing the uninformed to detector-informed settings. Public detection improves removal only at small edit budgets, since plain rephrasing already strips the watermark at a lower quality cost, but it does enable forgery, which the private pipeline can identify. Overall, releasing half of the watermark enables transparency and interoperability, and tampering with the released half stays detectable. This bounds the provider's liability and questions the need to keep detectors fully private.
Read the full abstract ▾
All authors Georgios Milis, Tom Sander, Tomáš Souček, Heng Huang, Pierre Fernandez
Analysis of PDEs
Invariant manifolds and transversality for a traveling-wave ODE in a controlled reaction-diffusion model
Stefano Bianchini and Chiara Trifone
We study the parameter dependence and transversality properties of trajectories of the controlled ODE $$ U_x = P, \qquad P_x = - f(U) - \beta P + \tilde \alpha(x)U, $$ arising from traveling-wave solutions of a bistable reaction-
diffusion equation with control. Optimal profiles minimizing the $L^1$-norm of the control $\tilde \alpha$ are determined by the unstable and stable
manifolds $\Gamma_u$, $\Gamma_s$, the critical curve $$ P^*(U) = \sqrt{Uf(U)}, $$ and the associated function $$ F(U) = \frac{dP^*}{dU} + \frac{f(U)}{P^*(U)}. $$ We establish generic transversality properties for these objects with respect to perturbations of the speed $\beta$ and the nonlinear source $f$, together with quantitative estimates on the first- and second-order dependence of the invariant
manifolds on these parameters. We also analyze the loss of regularity at the critical speed $\beta^{**}$, where the unstable
manifold approaches the strongly stable
manifold of the
equilibrium $(U^*,0)$. These results provide the ODE ingredients used in the analysis of the structure and regularity of optimal profiles in paper [S. Bianchini and C. Trifone. Traveling profiles and control cost for a PDE describing the
evolution of invasive species. 2026. arXiv: 2606.23979 [math.OC]].
Read the full abstract ▾
Computer Vision
VINCIE-NExT: Unlocking Video Editing from Images via In-Context Modeling
Leigang Qu, Feng Cheng, Ziyan Yang et al. (+7 more)
Building a capable
video editor remains significantly harder than a
video generator: editing requires (source, instruction, edited) triplets that are prohibitively expensive to annotate and difficult to synthesize at scale, whereas image editing has already reached maturity with millions of such pairs readily available. In this work, we introduce VINCIE-NExT, a unified framework that transfers editing capability from images to
videos through in-context visual demonstrations, alleviating the need for large-scale paired
video editing data. VINCIE-NExT decomposes
video editing into a structured chain of composable sub-tasks (
Video -> Image -> Image ->
Video), routing editing intent through the image domain and enabling scalable joint training from heterogeneous image and
video corpora under a unified
diffusion objective. An image editing pair, synthesized by the model or supplied by the user, is prepended as an in-context visual demonstration that serves as a spatial appearance blueprint for every output frame. To ground appearance edits across the interleaved context, we introduce a novel position encoding that links image demonstrations and
video frames in a shared spatial coordinate system, enabling pixel-faithful propagation of appearance changes to every output frame. Chain-of-Editing further provides principled test-time scaling: by executing the sub-task chain as progressive
diffusion stages, editing quality can be improved by investing additional compute without retraining. Comprehensive experiments on OpenVE-Bench demonstrate the state-of-the-art performance across diverse editing categories, with ablations confirming the effectiveness of each component.
Read the full abstract ▾
All authors Leigang Qu, Feng Cheng, Ziyan Yang, Bangbang Yang, Zhaoyang Huang, Wei Chow, Yicong Li, Wenjie Wang, Tat-Seng Chua, Yan Zeng
Systems and Control
Predefined-Time Integral Reinforcement Learning for Unknown Nonlinear Systems via Inverse-Optimal Design
Tien Dat Vu
This paper develops a new predefined-time integral reinforcement learning framework for optimal control of unknown nonlinear systems. The unknown drift is first approximated by a radial basis function (RBF) neural network, together with a data-driven online identification law for updating the corresponding neural weights. After the identifier converges to a sufficiently small neighborhood of the true dynamics, the learned model is incorporated into the integral reinforcement learning (IRL) problem. Unlike conventional reinforcement-learning-based optimal control, the desired convergence time is introduced directly into the control objective: a Lyapunov function and its prescribed decay behavior are specified by the designer, and inverse-optimal control is then used to construct a compatible running cost whose optimal policy inherits the predefined-time stabilization property. The value function is approximated by a second RBF neural network, and a new critic update law is developed to impose predefined-time convergence on the critic weights. Finite informative learning data are stored in a replay buffer and reused during the critic update, thereby avoiding the need for persistent excitation throughout the closed-loop operation. Theoretical analysis proves that the critic-weight error enters a prescribed residual set within the allocated learning horizon, while the closed-loop state reaches a small neighborhood of the origin within the overall designer-specified deadline. Numerical simulations on an unknown nonlinear system verify accurate drift reconstruction, predefined-time critic learning, and closed-loop convergence.
Read the full abstract ▾
Machine Learning
Few-Step Generation via Data-Space Iteration
Shanchuan Lin, Yansong Peng, Fu-Yun Wang et al. (+1 more)
Flow matching has emerged as a scalable paradigm for training high-quality generative models, but sampling from the learned probability flow requires many network evaluations.
Distillation can reduce this cost to one or a few evaluations; however, one-step generation often sacrifices quality, making few-step generation the practical operating regime. Existing few-step methods perform their iterative computation along the probability flow and therefore require a fixed, manually chosen timestep discretization. This discretization is often chosen heuristically and is expensive to tune; it may also be restrictive when refinement difficulty differs across samples or spatial locations. We introduce data-space iteration, a few-step generation framework that removes flow discretization altogether. Starting from noise, a shared generator directly refines its prediction in data space, with every iteration trained to produce the best sample permitted by its capacity. Our formulation integrates with distribution matching
distillation (DMD) with minimal changes, enabling a controlled comparison between iteration methods under matched training settings. On class-conditional ImageNet 256x256, data-space iteration outperforms standard discretization baselines and matches or improves upon variants selected through schedule search, without requiring schedule-specific training. These results show that data-space iteration provides a simple and effective alternative to discretized flow-space iteration for fast generation.
Read the full abstract ▾
All authors Shanchuan Lin, Yansong Peng, Fu-Yun Wang, Haoqi Fan
High Energy Astrophysical Phenomena
An updated model of neutrino emission from proto-neutron stars
Matteo Ballelli, Federica Pompa, Christoph Ternes et al. (+2 more)
We present an improved parametric model for
neutrino emission from core-collapse
supernovae that directly connects
neutrino luminosity and average energy to
stellar properties during emission. Our model incorporates two critical physical processes: convection within the proto-
neutron star and the
evolution of the neutrinosphere radius during the early emission phase. Using simulated
datasets of
supernova neutrino events in the Super-Kamiokande detector, we assess the sensitivity to astrophysical properties in the event of a future detection. We show that the
neutrino mass ordering plays a crucial role in determining which parameters can be constrained by the observed
neutrino data.
Read the full abstract ▾
All authors Matteo Ballelli, Federica Pompa, Christoph Ternes, Marco Drago, Giulia Pagliaroli
Cosmology and Nongalactic Astrophysics
A New Era for Supernova Science with the James Webb Space Telescope
Justin Pierel, David Coulter, Christa DeCoursey et al. (+17 more)
Supernovae (SNe) are excellent probes of the physical, chemical, and cosmological context of
galaxies across the history of the Universe. The James Webb Space
Telescope (JWST ) has a truly unique combination of wavelength, sensitivity, and resolution, providing a new window into SN physics, progenitors,
dust production,
redshift evolving
stellar populations, and cosmology since its launch. Here we summarize the astounding observations JWST has made of SNe in its first few years of operations, as well as the corresponding advancements in our understanding of the Universe. This includes detailed views of nearby SNe, the first spectroscopic samples of high-z SNe, and a significant number of gravitationally lensed SNe valuable for both cosmological inference and high signal-to-noise (S/N) spectroscopy. While the upcoming Nancy Grace Roman Space
Telescope is expected to revolutionize space-based time-domain astronomy, JWST and its unique capabilities will continue to play a complementary role by providing high-precision spectroscopic and photometric data for a broad range of SN science cases.
Read the full abstract ▾
All authors Justin Pierel, David Coulter, Christa DeCoursey, James DerKacy, Ori Fox, Melissa Shahbandeh, Matthew Siebert, Michael Engesser, Conor Larison, Thomas Moore, Armin Rest, Rodrigo Angulo, Michael Dulude, Eiichi Egami, Massimo Griggio, Mitchell Karmen, Xiaolong Li, Estefania Padilla Gonzalez, Koji Shukawa, Louis-Gregory Strolger
Robotics
UNITAS: A 3D-Native World Action Model for Embodied Manipulation
Ruixiang Wang, Yongyi Su, Wenlve Zhou et al. (+14 more)
World action models (WAMs) aim to answer a coupled physical question: given a task instruction, what motion should the
robot execute, and how will that motion change the surrounding world? Most existing WAMs build on pretrained
video generators and represent world
evolution through images or visual latents. Robotic interaction, however, takes place in metric three-dimensional space, while images are view-dependent projections whose pixel distances do not directly encode physical distances. We introduce UNITAS, to our knowledge the first 3D-native world action model that unifies observations, actions, and scene dynamics in a shared metric 3D frame within each interaction, using a common representation across
robot embodiments and human hands. Action flow represents human hands and
robot grippers as 3D point trajectories, while scene flow describes scene-point displacements conditioned on these trajectories. World-aligned 3D positional
embeddings ground visual
tokens with or without depth input, and a physical-time trajectory tokenizer encodes each point trajectory as one
token anchored at its current 3D position. This interface supports both direct action execution and action-conditioned scene prediction. With 1.7B parameters, UNITAS achieves the best action-conditioned scene prediction on RoboTwin among the compared methods, with up to 49% lower displacement errors than PointWorld, and state-of-the-art manipulation success, including 99.8% on LIBERO and an average of 85% across real-world tasks. The code is available at https://github.com/DexForce/UNITAS.
Read the full abstract ▾
All authors Ruixiang Wang, Yongyi Su, Wenlve Zhou, Bo Yue, Hengyan Liu, Dekun Lu, Yuxin Tian, Yihan Fang, Zerui Wu, Xing Hu, Jietao Chen, Yong Guo, Ziyan He, Junbin Yuan, Guiliang Liu, Xiaofen Xing, Kui Jia
Quantum Physics
Nearly Optimal T-Count for Symmetry-Based Quantum State Purification
Jialiang Tang, Junhong Nie, Keming He et al. (+2 more)
Protecting
quantum information from noise is essential for reliable
quantum computation and
quantum memories.
Quantum state purification addresses this challenge by processing multiple noisy copies of an unknown
quantum state to produce an output state of higher fidelity. Here we establish nearly tight upper and lower bounds on the non-Clifford resources required to implement the symmetric-projection purification protocol. For $n$ copies of a $d$-dimensional system, we construct a Clifford+T
circuit that block-encodes the symmetric projector with operator-norm error at most $\epsilon$ using $\widetilde{O}(n)$ T gates, where the tilde-$O$ notation suppresses polylogarithmic factors in $n$, $d$, and $1/\epsilon$. Conversely, using stabilizer nullity, we prove that any Clifford+T implementation of such a block encoding with error $\epsilon<\frac{1}{2}\sqrt{1-1/n}$ requires at least $2n-2$ T gates, even allowing arbitrary positive normalization and clean stabilizer ancillas. Taken together, these results establish the nearly optimal scaling of non-Clifford cost with the number of copies and uncover the fundamental resource requirements for coherent
symmetry projection in
quantum state purification.
Read the full abstract ▾
All authors Jialiang Tang, Junhong Nie, Keming He, Xin Wang, Xiaoming Sun