Search arXiv⌕ Search

arXiv subjects

Teng Wang

Publications and source records attributed to Teng Wang.

At least 37 records · Page 2Linked to original sources
arXiv2026

Ab initio lattice calculation of nuclear magnetic dipole moments with systematic error quantifications

Nuclear magnetic moments are sensitive probes of nuclear structure. However, their accurate quantitative description poses significant challenges, demanding both accurate nuclear and electromagnetic interactions as well as rigorous control of algorithmic uncertainties. Here, we present the first systematic calculation of magnetic dipole moments for selected light nuclei and aluminum isotopes within nuclear lattice effective field theory (NLEFT), an \textit{ab initio} framework applicable to medium-mass and heavy nuclei. Our calculations employ a lattice next-to-next-to-next-to-leading-order (N$^3$LO) chiral interaction together with electromagnetic currents consistently derived up to the two-body level. To achieve controlled predictions, we incorporate recently developed NLEFT algorithms and perform a comprehensive assessment of algorithmic uncertainties. Within the estimated uncertainties, our results are in good overall agreement with experiment and demonstrate that two-body currents are essential for reproducing the observed magnetic moments. We further benchmark our predictions against other \textit{ab initio} calculations for light nuclei ($A\leq12$). Our work establishes a solid foundation for \textit{ab initio} studies of electroweak observables using methods that scale efficiently to medium-mass and heavy nuclei while demonstrating state-of-the-art accuracy.

nucl-th↗
arXiv2026

Isoparametric foliations of $S^n\times S^n$ of codimension $1$

In this article, we give several families of new examples of isoparametric foliations of $S^n\times S^n$; establish some basic properties of isoparametric foliations of $S^n\times S^n$ of codimension $1$, especially any of them can be realized as level sets of the restriction of a bi-homogeneous polynomial of bi-degree $(1,1),(2,2),(3,3),(4,4)$ or $(6,6)$ to $S^n\times S^n$; and classify bi-homogeneous isoparametric polynomials of bi-degree $(1,1)$ and $(2,2)$ and isoparametric foliations on $S^n\times S^n$ with $n$ even and not equal to $4$.

math.DG↗
arXiv2026

The anomalous long lifetime of $^{14}$C revealed by ab initio nuclear lattice EFT

The 5730-year half-life of $^{14}$C, the physical basis of radiocarbon dating, is anomalously long compared to typical nuclear-physics expectations. Its origin has remained a subject of debate for many decades. Here we report an \textit{ab initio} nuclear lattice effective field theory (NLEFT) calculation of $^{14}$C $β$ decay. Using systematically optimized interactions and transition operators consistently derived from chiral effective field theory, We obtain a result consistent with the Gamow-Teller matrix element $M_\text{GT}^\text{exp}\simeq 2\times 10^{-3}$ measured with the current uncertainty of $O(10^{-2})$. We first show that chiral interactions and weak currents beyond leading-order are essential for quenching the GT matrix element to the physical value, among which the optimization of three-nucleon forces is indispensable. We then illustrate that the quenching is deeply rooted in the ground-state structure of $^{14}$N as found in the nuclear shell model, where the competition between $S$- and $D$-wave components exists, sensitive to the interaction employed. The physical $^{14}$N ground state is found to be dominated by $D$-wave configurations, which constitutes the key factor for the quenching. The sensitivity of the decay matrix element to the fine-tuning of low-energy-constants is explored, revealing the prominent role of the $^3S_1$-channel two-nucleon contact force and the one-pion-exchange three-nucleon force. This work eliminates the gap between shell-model and \textit{ab initio} studies of $^{14}$C $β$ decay, provides a valid and straightforward explanation for the anomalous long lifetime of $^{14}$C, and turns NLEFT into a practical tool for the systematic study of nuclear transitions.

nucl-th↗
arXiv2026

Before Thinking, Learn to Decide: Proactive Routing for Efficient Visual Reasoning

Large multimodal models have achieved strong reasoning on complex visual tasks, but their inference efficiency is often restricted by long chains of thought. A promising solution is to pair a small draft model with a large target model, enabling cooperative inference employing a routing signal that adaptively routes queries to either the draft or target model based on their difficulties for optimal efficiency and accuracy. Yet, the remaining bottleneck is to establish a reliable query difficulty signal under multimodal settings. Existing approaches designed for language models either rely on post-hoc token probabilities, which fall short in multimodal scenarios, or depend on supervised fine-tuning, which is a data-sensitive strategy. Both paradigms perform routing only after a complete output, and ignore whether the target model can actually solve the routed instances. To address this, we propose PRP, a Proactive Routing Paradigm that enables early decision-making by jointly evaluating the competence of both the draft and target models. Our Draft Rating Learning (DRL) equips the draft model with an internal confidence estimator, while Joint Rating Learning (JRL) predicts how well the target model can handle a given query, thereby prioritizing the allocation of samples it excels at rather than the hardest ones. These ratings enable fine-grained, instance-level \textbf{Proactive Routing} and substantially accelerate inference without compromising overall performance. Extensive experiments across multiple multimodal reasoning benchmarks validate our effectiveness and efficiency.

cs.CL↗
arXiv2026

FlowCIR: Semantic Transport via Flow Matching for Zero-Shot Composed Image Retrieval

Zero-shot composed image retrieval (ZS-CIR) aims to retrieve a target image by editing a reference image with a natural-language instruction, without relying on domain-specific annotated triplets. Most existing ZS-CIR methods rely on textual inversion to translate the reference image into pseudo-text tokens and then compose them with the instruction via simple concatenation in the text space, which can be lossy and brittle for fine-grained semantics. In this work, we propose a new paradigm, namely FlowCIR, that casts ZS-CIR as conditional semantic transport between reference and target embeddings. Leveraging \emph{conditional flow matching}, our model learns a lightweight transport field that maps the instruction representation toward a target-aligned query embedding conditioned on the reference image. Since FlowCIR operates on pre-extracted VLM embeddings and trains only a small transport module without updating the image or text encoder, it offers a computationally efficient training protocol compared with prior textual-inversion-based approaches. The resulting framework is training-efficient, requiring roughly $10\times$ fewer training resources than prior textual-inversion-based approaches. We further identify negation and removal as a major failure mode of VLM-based composition. To address this, we propose an inference-only Multi-Negative Steering strategy that steers a negation-containing relative instruction away from its negated semantics, mitigating the limited negation handling of VLMs and improving robustness on negation-heavy queries. Extensive experiments on standard CIR benchmarks demonstrate that FlowCIR achieves strong and competitive performance compared with recent ZS-CIR methods.

cs.CV↗
arXiv2026

UltraStar: Semantic-Aware Star Graph Modeling for Echocardiography Navigation

Echocardiography is critical for diagnosing cardiovascular diseases, yet the shortage of skilled sonographers hinders timely patient care, due to high operational difficulties. Consequently, research on automated probe navigation has significant clinical potential. To achieve robust navigation, it is essential to leverage historical scanning information, mimicking how experts rely on past feedback to adjust subsequent maneuvers. Practical scanning data collected from sonographers typically consists of noisy trajectories inherently generated through trial-and-error exploration. However, existing methods typically model this history as a sequential chain, forcing models to overfit these noisy paths, leading to performance degradation on long sequences. In this paper, we propose UltraStar, which reformulates probe navigation from path regression to anchor-based global localization. By establishing a Star Graph, UltraStar treats historical keyframes as spatial anchors connected directly to the current view, explicitly modeling geometric constraints for precise positioning. We further enhance the Star Graph with a semantic-aware sampling strategy that actively selects the representative landmarks from massive history logs, reducing redundancy for accurate anchoring. Extensive experiments on a dataset with over 1.31 million samples demonstrate that UltraStar outperforms baselines and scales better with longer input lengths, revealing a more effective topology for history modeling under noisy exploration. Code is available at https://github.com/LeapLabTHU/UltraStar.

cs.CV↗
arXiv2026

CycleRL: Sim-to-Real Deep Reinforcement Learning for Robust Autonomous Bicycle Control

Autonomous bicycles offer a promising agile solution for urban mobility and last-mile logistics. However, conventional control strategies often struggle with underactuated nonlinear dynamics, suffering from sensitivity to model mismatches and limited adaptability to real-world uncertainties. To address this, we develop CycleRL, a comprehensive sim-to-real framework for robust autonomous bicycle control. Our approach establishes a direct perception-to-action mapping within the high-fidelity NVIDIA Isaac Sim environment, leveraging Proximal Policy Optimization (PPO) to optimize the control policy. The framework features a composite reward function tailored for concurrent balance maintenance, velocity tracking, and steering control. Crucially, systematic domain randomization is employed to reduce the reliance on precise system modeling, bridge the simulation-to-reality gap and facilitate direct transfer. In simulation, CycleRL achieves promising performance, including a 99.90% balance success rate, a heading tracking error of 1.15°, and a velocity tracking error of 0.18 m/s. These quantitative results, coupled with successful hardware deployment, validate DRL as an effective paradigm for autonomous bicycle control, offering superior adaptability over traditional methods. Video demonstrations are available at https://cpnt-lab.github.io/CycleRL/.

cs.RO↗
arXiv2026

On the relationship between block spaces and Orlicz spaces

Let $1 -1$. Let $(X,d,μ)$ be an $s$-Ahlfors-regular quasi-metric measure space. Suppose that $B^{0,v}_q(X)$ is the block space which consists of all functions that admit a decomposition into $q$-blocks supported on balls. In this paper, we study the relationship between the block space $B^{0,v}_q(X)$ and the Orlicz-type space $L(\log^+\!\!L)^{1+v}(X)$. More precisely, we show that the block space $B_q^{0,v}(X)$ is a proper subspace of the Orlicz space $L(\log^+\!\!L)^{1+v}(X)$ for any fixed $1 -1$. Namely, $$B_q^{0,v}(X)\subsetneq L(\log^+\!\!L)^{1+v}(X),$$ which gives a confirmed answer to a longstanding open problem concerning the relationship between block spaces and Orlicz-type spaces on the unit sphere $\mathbb S^{n-1}$. We further show that $L(\log^+\!\!L)^{1+v}(X)$ is the smallest Orlicz-type space containing $B^{0,v}_{q}(X)$. We also introduce a generalized block space $\mathscr B_q^{0,v}(X)$ that depends only on the measure structure and show that this space is equivalent to the Orlicz space $L(\log^+\!\!L)^{1+v}(X)$ when $μ(X)<\infty$. Finally, we consider two special cases that further clarify the roles of the parameter $q$ and the logarithmic weight.

math.CA↗
arXiv2026

DeepImageSearch: Benchmarking Multimodal Agents for Context-Aware Image Retrieval in Visual Histories

Existing multimodal retrieval systems excel at semantic matching but implicitly assume that query-image relevance can be measured in isolation. This paradigm overlooks the rich dependencies inherent in realistic visual streams, where information is distributed across temporal sequences rather than confined to single snapshots. To bridge this gap, we introduce DeepImageSearch, a novel agentic paradigm that reformulates image retrieval as an autonomous exploration task. Models must plan and perform multi-step reasoning over raw visual histories to locate targets based on implicit contextual cues. We construct DISBench, a challenging benchmark built on interconnected visual data. To address the scalability challenge of creating context-dependent queries, we propose a human-model collaborative pipeline that employs vision-language models to mine latent spatiotemporal associations, effectively offloading intensive context discovery before human verification. Furthermore, we build a robust baseline using a modular agent framework equipped with fine-grained tools and a dual-memory system for long-horizon navigation. Extensive experiments demonstrate that DISBench poses significant challenges to state-of-the-art models, highlighting the necessity of incorporating agentic reasoning into next-generation retrieval systems.

cs.CV↗
arXiv2026

VA-Adapter: Adapting Ultrasound Foundation Model to Echocardiography Probe Guidance

Echocardiography is a critical tool for detecting heart diseases, yet its steep operational difficulty causes a shortage of skilled personnel. Probe guidance systems, which assist in acquiring high-quality images, offer a promising solution to lower this operational barrier. However, robust probe guidance remains challenging due to significant individual variability. This variability manifests as differences in low-level features within two-dimensional (2D) images, which complicates image feature understanding, and differences in individual three-dimensional (3D) structures, which poses challenges for precise navigation. To address these challenges, we first propose leveraging the robust image representations learned by ultrasound foundation models from vast datasets. Yet, applying these models to probe navigation is non-trivial due to their lack of understanding of individual 3D structures. To this end, we meticulously design a Vision-Action Adapter (VA-Adapter) to online inject the capability of understanding individual 3D structures. Specifically, by embedding the VA-Adapter into the foundation model's image encoder, the model can infer cardiac anatomy from historical vision-action sequences, mimicking the cognitive process of a sonographer. Extensive experiments on a dataset with over 1.31M samples demonstrate that the VA-Adapter outperforms strong probe guidance models while requiring approximately 33 times fewer trained parameters. Code is available at https://github.com/LeapLabTHU/VA-Adapter.

cs.CV↗
arXiv2026

Chain-of-Glimpse: Search-Guided Progressive Object-Grounded Reasoning for Video Understanding

Video understanding requires identifying and reasoning over semantically discriminative visual objects across frames, yet existing object-agnostic solutions struggle to effectively handle substantial object variations over time. To address this, we introduce Chain-of-Glimpse, a search-guided progressive object-grounded reasoning framework that explicitly anchors each reasoning step to specific visual evidence regions, enabling compositional and multi-step decision-making. Formally, Chain-of-Glimpse formulates video reasoning as a step-by-step process that incrementally builds spatially grounded traces around task-relevant visual objects, thereby mitigating over-reliance on saliency-driven cues. Specifically, Chain-of-Glimpse features a search-guided controller, optimized via reinforcement learning with a format reward that significantly incentivizes grounding capability, to iteratively ground visual evidence regions and form reliable reasoning trajectories, yielding accurate and interpretable multi-step decisions. Extensive evaluations on both in domain NExTQA and out-of-domain Video-Holmes, CG-Bench Reasoning, and VRBench benchmarks demonstrate consistent performance gains, robustness and generalization of Chain-of-Glimpse across diverse video reasoning tasks.

cs.CV↗
arXiv2026

Nuclear charge radii of aluminium isotopes at the proton drip line

Understanding the evolution of nuclear size away from stability remains a central challenge in nuclear physics. In neutron-deficient systems, charge radii can be highly sensitive to the interplay between strong and electromagnetic interactions, and the effects of weak binding, giving rise to exotic nuclear phenomena. However, experimental data on these systems has been limited by short lifetimes and low production rates. Here we report the first laser-spectroscopy measurements of nuclear charge radii along the neutron-deficient aluminium isotopic chain, from $^{25}$Al to the proton-drip-line nucleus $^{22}$Al, using the {Resonance Ionization Spectroscopy Experiment} (RISE) at the {Facility for Rare Isotope Beams} (FRIB). Our measurements reveal a step-like increase in charge radius toward the drip line, with similar radii for $^{22,\,23}$Al. A comparison of our results with those of their mirror partners reveals an almost identical correlation with the calculated proton skins and is consistent with the systematic trend of well-bound nuclei. These results offer insight for understanding the evolution of nuclear size at the proton dripline and place important constraints on modern nuclear theory. They also demonstrate the unique combined capabilities of RISE and FRIB to probe the structures of previously inaccessible nuclei at the limits of existence.

nucl-ex↗
arXiv2026

ImagineNav++: Prompting Vision-Language Models as Embodied Navigator through Scene Imagination

Visual navigation is a fundamental capability for autonomous home-assistance robots, enabling long-horizon tasks such as object search. While recent methods have leveraged Large Language Models (LLMs) to incorporate commonsense reasoning and improve exploration efficiency, their planning remains constrained by textual representations, which cannot adequately capture spatial occupancy or scene geometry--critical factors for navigation decisions. We explore whether Vision-Language Models (VLMs) can achieve mapless visual navigation using only onboard RGB/RGB-D streams, unlocking their potential for spatial perception and planning. We achieve this through an imagination-powered navigation framework, ImagineNav++, which imagines future observation images from candidate robot views and translates navigation planning into a simple best-view image selection problem for VLMs. First, a future-view imagination module distills human navigation preferences to generate semantically meaningful viewpoints with high exploration potential. These imagined views then serve as visual prompts for the VLM to identify the most informative viewpoint. To maintain spatial consistency, we develop a selective foveation memory mechanism, which hierarchically integrates keyframe observations via a sparse-to-dense framework, constructing a compact yet comprehensive memory for long-term spatial reasoning. This approach transforms goal-oriented navigation into a series of tractable point-goal navigation tasks. Extensive experiments on open-vocabulary object and instance navigation benchmarks show that ImagineNav++ achieves SOTA performance in mapless settings, even surpassing most map-based methods, highlighting the importance of scene imagination and memory in VLM-based spatial reasoning.

cs.RO↗
arXiv2026

ColorBrowserAgent: Complex Long-Horizon Browser Agent with Adaptive Knowledge Evolution

With the advancement of vision-language models, web automation has made significant progress. However, deploying autonomous agents in real-world settings remains challenging, primarily due to site heterogeneity, where generalist models lack domain-specific priors for diverse interfaces, and long-horizon instability, characterized by the accumulation of decision drift over extended interactions. To address these challenges, we introduce ColorBrowserAgent (Complex Long-Horizon Browser Agent), a knowledge-evolving agent for robust web automation. Our approach addresses these challenges through two synergistic mechanisms: human-in-the-loop knowledge adaptation that transforms sparse human feedback into reusable domain knowledge, and knowledge-aligned progressive summarization that stabilizes long interactions through memory compression. Extensive experiments on WebArena, WebChoreArena and industrial deployment show that ColorBrowserAgent consistently outperforms strong baselines. It achieves a state-of-the-art success rate of 71.2% on WebArena and maintains 47.4% performance under zero-shot transfer setting on WebChoreArena. In commercial deployment, it improves user satisfaction by 19.3% relatively, verifying its robustness in real-world scenarios.

cs.HC↗
arXiv2026

LiveVLN: Breaking the Stop-and-Go Loop in Vision-Language Navigation

Recent navigation systems achieve strong benchmark results, yet real-world deployment often remains visibly stop-and-go. This bottleneck arises because the sense-inference-execution loop is still blocking: after each new observation, the controller must wait for sensing, transmission, and inference before motion can continue. Reducing action-generation cost alone therefore does not remove redundant waiting. To address this issue, we present LiveVLN, a training-free framework for more continuous embodied navigation by augmenting pretrained VLM navigators with multi-step action continuation. Instead of pausing for each full sense-and-inference round, LiveVLN overlaps execution with the processing of newly arrived observations, allowing refreshed future actions to be handed off before the current executable prefix is exhausted. This design keeps actions continuously available during motion, reducing idle waiting and enabling smoother online execution. The framework operates at runtime and can be integrated with compatible pretrained VLM navigators. Across R2R and RxR, LiveVLN preserves benchmark performance while reducing waiting time and improving action availability. In real-world deployments, it cuts average episode waiting time by up to $77.7\%$ and shortens wall-clock episode time by $12.6\%$ on StreamVLN and $19.6\%$ on NaVIDA, yielding more coherent execution during deployment. Code is available at https://github.com/NIneeeeeem/LiveVLN.

cs.RO↗
arXiv2026

OmniScript: Towards Audio-Visual Script Generation for Long-Form Cinematic Video

Current multimodal large language models (MLLMs) have demonstrated remarkable capabilities in short-form video understanding, yet translating long-form cinematic videos into detailed, temporally grounded scripts remains a significant challenge. This paper introduces the novel video-to-script (V2S) task, aiming to generate hierarchical, scene-by-scene scripts encompassing character actions, dialogues, expressions, and audio cues. To facilitate this, we construct a first-of-its-kind human-annotated benchmark and propose a temporally-aware hierarchical evaluation framework. Furthermore, we present OmniScript, an 8B-parameter omni-modal (audio-visual) language model tailored for long-form narrative comprehension. OmniScript is trained via a progressive pipeline that leverages chain-of-thought supervised fine-tuning for plot and character reasoning, followed by reinforcement learning using temporally segmented rewards. Extensive experiments demonstrate that despite its parameter efficiency, OmniScript significantly outperforms larger open-source models and achieves performance comparable to state-of-the-art proprietary models, including Gemini 3-Pro, in both temporal localization and multi-field semantic accuracy.

cs.CV↗
arXiv2026

Scheduling Cause-Effect Chains without Timing Anomalies in End-to-End Latency

In real-time systems, both individual task execution and data propagation must meet strict timing constraints. Cause-effect (CE) chains are widely used to analyze such behaviors by end-to-end latency. However, timing anomalies (TAs) can distort it, where a local reduction in execution times leads to an increase in the overall end-to-end latency. As a result, precisely analyzing the upper bounds of the latency becomes challenging, and such systems typically exhibit larger upper bounds than TA-eliminated systems. Existing studies either eliminate TAs by completely sacrificing average latency to simplify analysis or, despite adopting complex safe analysis methods, do not eliminate TAs effectively, still having high latencies. To address this issue, we identify two basic causes of TAs in end-to-end latency. Based on these causes, we propose the first treatment that eliminates TAs in the latency with negligible average latency loss using Deterministic Data Flow (DDF). We further formally prove its TA-free property. Therefore, we can get a precise upper bound for latency when all jobs execute with their worst-case execution times. Experimental results show that it effectively reduces the maximum end-to-end latency, the average latency, and latency jitter compared with the state-of-the-art (SOTA) method.

eess.SY↗
arXiv2026

Towards Hierarchical Multi-Step Reward Models for Enhanced Reasoning in Large Language Models

Recent studies show that Large Language Models (LLMs) achieve strong reasoning capabilities through supervised fine-tuning or reinforcement learning. However, a key approach, the Process Reward Model (PRM), suffers from reward hacking, making it unreliable in identifying the best intermediate step. In addition, the cost of annotating reasoning processes for reward modeling is high, making large-scale collection of high-quality data challenging. To address this, we propose a novel reward model approach called the Hierarchical Reward Model (HRM), which evaluates both individual and consecutive reasoning steps at both fine-grained and coarse-grained levels. HRM excels at assessing multi-step reasoning coherence, especially when flawed steps are later corrected through self-reflection. To further reduce the cost of generating training data, we introduce a lightweight and effective data augmentation strategy called Hierarchical Node Compression (HNC), which merges two consecutive reasoning steps into one within the tree structure. By applying HNC to MCTS-generated reasoning trajectories, we enhance the diversity and robustness of HRM training data while introducing controlled noise with minimal computational overhead. Empirical results on the PRM800K dataset show that HRM, together with HNC, provides more stable and reliable evaluations than PRM. Furthermore, cross-domain evaluations on the MATH500 and GSM8K datasets demonstrate HRM's strong generalization and robustness across a variety of reasoning tasks.

cs.CL↗