Chemical Physics
Papers
Fresh from arXiv, typeset properly. Highlighted words trade on the market.
All of arXiv
Thursday 27 August
Neural-network quantum states (NNQSs) can represent many-electron wave functions without explicitly enumerating the determinant space, but their accuracy depends jointly on model size and variational-optimization effort. Here we characterize this dependence for a physics-conditioned autoregressive NNQS trained separately on two six-molecule source benchmarks. Across eight model sizes and five optimization milestones, we find that model size and optimization steps jointly shape the energy error. The capacity advantage of larger models becomes more apparent with sufficient optimization, while the returns from additional optimization vary with model size. We capture this coupling using an interaction scaling law and quantify the cumulative compute of each evaluated configuration. The resulting error-compute Pareto frontiers provide a practical decision rule for jointly selecting model size and optimization steps under a given compute budget within the evaluated range. Furthermore, we find that this beneficial scaling trend persists during fine-tuning on held-out N$_2$. Pretrained models show decreasing error with increasing model size, with a steeper reduction following pretraining on the Hard benchmark. Together, these results place autoregressive neural quantum states within the broader landscape of empirical neural scaling and open a quantitative route toward the systematic scaling of neural quantum solvers for ab initio quantum chemistry.
Social and Information Networks
A Novel Information Workflow for Structural Behavioural Analysis in Dynamic Attributed Graphs
Social systems play a vital role through interactions around shared interests and the maintenance of relationships, which are part of society. These can be represented as networks, and their analysis plays a crucial role in dynamic environments, which can be achieved through link prediction and community detection tasks. Several link prediction approaches estimate new relationships, such as suggesting friends on social media. On the other hand, a social faction is a group of connected people that reveals the internal structure of the social system and can be identified through community detection approaches. The evolution of social systems can be understood using these approaches by analysing their structure and behaviour. To handle this, a combination of these approaches, along with interaction-based analysis, is used to form cohesive communities, thereby enabling the study of social systems dynamically. In addition, attribute information plays a critical role. Most real-world graphs, like Facebook and Twitter, have attribute information that provides context, and integrating them with structural information offers deeper insights into patterns, which is a critical task. However, existing approaches have limitations in handling dynamics of social systems. Additionally, no existing approaches in the literature help identify potential links and form densely connected communities for attributed graphs. To address these challenges, we have proposed an information workflow, i.e., inc-LPCDAG (\underline{inc}remental Link Prediction and Community Detection in Attributed Graphs). We conducted experiments to evaluate efficiency, and results demonstrate the effectiveness of our proposed workflow by analysing and understanding structural and behavioural processes.
Read the full abstract
Quantum Physics
Anticoncentration and entanglement in Gaussian boson sampling
Anticoncentration and entanglement are two properties used to study classical hardness or easiness of Gaussian boson sampling in varying regimes. In this paper, we consider three directions concerning anticoncentration and entanglement in Gaussian boson sampling: (1) We derive closed-form expressions for the second moment of hafnians of symmetric Gaussian products, and use this to precisely locate the anticoncentration transition as a function of the number of squeezed input modes. (2) We derive closed-form expressions for the Rényi-$\alpha$ Page curves for Gaussian boson sampling. (3) We study unequal input squeezing parameters $(s_i)_i$, and prove estimates as well as monotonicity of the average-case Rényi-2 entropy Page curve in terms of the magnitudes $|s_i|$, which allow for extending equal squeezing results to unequal squeezing.
Read the full abstract
Computation and Language
Topic Matching in the Wild: Benchmark and Lessons from Real-World ASR Transcripts
In contact centers, real-time agent-assist tools determine, for each of many predefined topics, whether a live customer utterance is relevant and display a coaching card to the agent when it is. The input is noisy and challenging: ASR(Automatic Speech Recognition) transcripts of spontaneous phone conversations, which can be unclear, repetitive, and mostly lack punctuation. To systematically study this real-world task, we curate a human-annotated topic-utterance judgments dataset sourced from real call-center transcripts. We compare three types of matchers: a regex-based baseline, zero-shot sentence embedding encoders, and Gemini-based LLM matchers. In addition, two types of topic representations are studied in our benchmark:keyphrases and natural language description. Our empirical experiments highlight the superior performance of lightweight LLM matchers over embedding and regex models when equipped with natural language descriptions.
Read the full abstract
Computation and Language
The Curse of Multilinguality in Lexical Normalization
Lexical normalization rewrites the noisy, non-standard words that fill user-generated text (tmrw, u, gr8) into their standard forms. Because labelled data is scarce for most languages, a popular shortcut is to train a single model on many languages at once. We ask a simple question: how many languages should such a model be trained on? Using one fixed-capacity character-level model and twelve languages from a standard benchmark, we vary the number of jointly trained languages from one to twelve and measure per-language accuracy. We find a clear curse of multilinguality: accuracy is highest when a language is trained with only a few others, often just one to four, and then falls steadily and substantially, dropping by about forty percent as the rest are piled on. A control that holds the total amount of training data constant makes the decline arrive sooner and fall further, which points to competition among the languages for one fixed-size model rather than to how much data is available. We also test whether a language's typological distance from the others predicts its ideal number of co-training languages, and find no dependable rule: any apparent relationship rests on a couple of languages and does not hold up. For compact normalization models, less can be more: a few languages beat pooling everything into a single model.
Read the full abstract
Artificial Intelligence
UI-Venus-2 Technical Report
Multimodal GUI agents have emerged as a promising paradigm for digital task automation, yet transitioning from benchmark-oriented models to dependable real-world applications remains challenging due to limited environment coverage, brittle task construction, and unreliable reward verification. In this work, we present UI-Venus-2, a general-purpose foundation GUI agent designed to operate across mobile, web, and desktop environments through a unified closed-loop reasoning-action framework. To bridge the gap toward practical deployment, we jointly scale three critical dimensions: (1) Environments, expanding coverage to more than 170 multilingual mobile apps and native desktop operating systems; (2) Tasks, employing a deep-research pipeline for function-grounded instruction generation; and (3) Verification, adopting trace-level and sample-level evaluators with visual keypoints and multi-model voting to ensure reliable RL signals for training. Furthermore, we integrate safety-aware mechanisms to ensure controlled execution of consequential actions. By offering a capable, efficient, and open-source foundation, UI-Venus-2 advances the field toward more generalizable, verifiable, and self-reflective agents for real-world applications.
Read the full abstract
General Physics
Exploring thermodynamic and photonic properties of new black hole solutions in F(R) gravity theory
We investigate black hole solutions in F(R) gravity within a four-dimensional spacetime, separately treating the cases of vanishing and nonvanishing energy momentum tensors. We demonstrate that in the absence of an energy momentum tensor, the number of independent equations is sufficient to obtain the Ricci flat solutions without the need for a prescribed F(R) function. In the same case, also the solutions with variable Ricci scalar can obtained too. Furthermore, for a nonvanishing energy momentum tensor, we derive exact solutions for both constant and variable Ricci scalars, thereby introducing new charged black holes. Utilizing this new method, we not only recover the general relativistic and previously known F(R) gravity solutions but also introduce four new classes of F(R) black holes. We calculate their thermodynamic quantities and confirm the validity of the first law of black hole thermodynamics. The thermal stability of these new black holes is analyzed using both thermodynamic and geometrical approaches. Finally, the photon spheres and black hole shadows are studied within the F(R) gravity framework.
Read the full abstract
Neural and Evolutionary Computing
Benchmarking spiking neural networks across sensing modalities on edge devices
Edge computing systems need to support diverse sensing workloads under tight energy and memory constraints, thereby motivating deployment-aware model selection. Spiking neural networks (SNNs) are a promising alternative to conventional artificial neural networks (ANNs), yet systematic evidence for when and why they provide practical advantages remains limited. Here, we present a benchmark of SNNs across five sensing modalities and multiple edge devices, systematically evaluating spike encoding, neuron models, and network topologies under consistent training and deployment protocols. We find that SNN advantages are strongly modality-dependent: while SNNs achieve performance broadly comparable to ANNs across most workloads, wireless sensing emerges as a particularly favorable domain. Frequency-domain and feature-space analyses further explain this result by showing that spiking dynamics naturally align with the spectral-temporal structure of wireless signals. Our deployment analysis further shows that SNN advantages are not one-dimensional, with energy gains often accompanied by modality-dependent system costs. Finally, we provide an open-source framework for reproducible benchmarking and deployment profiling, offering a practical foundation for algorithm-software-hardware co-design on emerging edge and neuromorphic computing platforms.
Read the full abstract
Image and Video Processing
Enhancing MRI Brain Tumor Edge Detection: A Hybrid Preprocessing Approach Utilizing CLAHE
Accurate boundary delineation of brain tumors in Magnetic Resonance Imaging (MRI) is a critical yet formidable challenge in neuro-oncology due to inherent scanner noise, complex anatomical structures, and uneven illumination. Traditional edge detection algorithms, while computationally lightweight and mathematically interpretable, frequently fail to capture the diffuse, localized boundaries of edema when relying solely on global preprocessing and manual parameter tuning. To overcome these limitations, we propose a hybrid automated edge detection pipeline. Our approach integrates an optimally configured Contrast-Limited Adaptive Histogram Equalization (CLAHE) layer into a comprehensive morphological preprocessing framework, followed by a deterministic sequential parameter sweep to fully automate threshold selection. The proposed hybrid model demonstrated enhancement in detecting critical anatomical structures in a publicly available benchmark database from Kaggle. By intelligently amplifying localized gradients without overwhelming the image with background noise, our method achieved higher Recall (Sensitivity). Consequently, the overall F1-Score elevated, and the Structural Similarity Index (SSIM) improved, all while maintaining a highly efficient execution. This establishes our optimized pipeline as a highly practical and near real-time operational model for clinical diagnostics, offering a compelling alternative to computationally heavy deep learning approaches.
Read the full abstract
Software Engineering
STEP: A Modular Silent Trial Engine for Operational Evaluation of Digital Pathology AI in Routine Workflow
Prospective silent trials provide an important bridge between retrospective validation of artificial intelligence (AI) models and their use in clinical care by evaluating model performance and operational reliability on live clinical data without influencing patient management. In computational pathology, conducting silent trials requires integration across laboratory information systems, digital pathology infrastructure, computational resources, and model inference pipelines, and these workflows are often implemented using application-specific software. We developed the Silent Trial Engine for Pathology (STEP), a reusable software platform for orchestrating prospective silent trials of computational pathology AI models across heterogeneous clinical and computational environments. STEP separates common trial orchestration from institution-specific data access and compute infrastructure through modular adapter interfaces. The platform supports scheduled case discovery, per-slide inference submission, deterministic idempotency, failure recovery, result and ancillary-data ingestion, persistent trial and run state, and audit logging, with compute adapters supporting local execution and high-performance computing environments using LSF and Slurm. STEP was deployed at three institutions to support prospective silent evaluation of EAGLE, an AI model for predicting EGFR mutation status from hematoxylin and eosin-stained whole-slide images. By separating trial-level workflow logic from site-specific integrations, STEP enables a common execution framework to operate across heterogeneous pathology environments while maintaining durable and auditable trial state. This approach may reduce duplicated engineering effort and facilitate systematic real-world evaluation of computational pathology AI before interventional clinical deployment.
Read the full abstract
Computation and Language
ReVA: A Region-Aware Visual Assistant for Visually Grounded Question Answering
Multimodal Large Language Models (MLLMs) have achieved remarkable progress in Visual Question Answering (VQA), yet they continue to struggle with questions requiring precise spatial reasoning and fine-grained visual understanding. These limitations often manifest as object, attribute, and spatial hallucinations, where models generate confident but visually unsupported responses due to insufficient region-level and fine-grained visual grounding. To address this challenge, we propose ReVA, a region-aware VQA model that employs a frozen CLIP ViT-L/14 Vision Transformer (ViT) and a Qwen2.5-7B-Instruct large language model (LLM) connected through a dual bridge that aligns both whole-image and region-level representations with the LLM's embedding space. The image bridge maps final transformer block features into image tokens. The region bridge maps cropped features from enriched intermediate features across ViT blocks so early texture and later object cues are more evident, into K region tokens for every bounding box. ReVA uses a detector stack that supplies automatic zero-shot bounding boxes that are both question-agnostic and question-dependent, using RAM++ (Recognize Anything Model), spaCy, and Grounding DINO. The image tokens and region tokens are concatenated as an LLM prompt prefix to jointly encode scene-level context and fine-grained regional evidence when answering questions. Evaluated on VQAv2, MMBench, POPE, and SEED-Bench, ReVA achieves 82.85% mean F1 on POPE, compared with 81.14% for an image-token baseline without region tokens. These results demonstrate that explicit region-aware visual representations reduce object hallucination and improve the factual grounding of MLLMs.
Read the full abstract
Computer Vision
Variable-Granularity Tokenization for High-Resolution Object Detection
ViT detectors fix a uniform token grid before any learned stage. A native-resolution aerial detector must then choose between resolving few-pixel objects and staying inside compute and memory limits. We introduce VGTok, a training-free tokenizer that sets patch granularity per region from pixels, ahead of the encoder. VGTok scores each region by multi-scale morphological top-hat separability from its surround, then thresholds those scores at a per-image percentile, which fixes the token budget. A structure-tensor gate ($\lambda_{\min}$) refines only where two-dimensional object structure supports it, leaving one-dimensional clutter coarse. The resulting token set is a strict partition of the image. In a Co-DETR detector with an EVA-02 ViT-L encoder, VGTok clears every published VisDrone-val AP and AP$_S$ at every budget from 40% to 100% of tokens. At 40% it records 44.22 AP with three fifths of the sequence discarded before the first transformer block; dense, it reaches 48.38 AP, $6.08$ above the strongest published entry. VGTok transfers to AI-TOD-v2 untouched, same scorer and same rank, and sets a new state of the art at 37.27 AP and 19.51 AP$_{vt}$. As a pure drop-in into a frozen checkpoint it reaches 36.29 AP at 78.5% of tokens, above every published entry, where our 376.3M-parameter detector clears a 3.0B multi-expert model. We show that a token budget fixed before the backbone, from local separability and structure geometry alone, holds accuracy on the tiny-object regimes that dominate aerial detection, at $3.1\times$ less encoder compute and $1.9\times$ less encoder memory. Code and models are available at \href{https://github.com/khayrulbuet13/vgtok}{github.com/khayrulbuet13/vgtok} and \href{https://huggingface.co/khayrulbuet13/vgtok}{huggingface.co/khayrulbuet13/vgtok}.
Read the full abstract
Image and Video Processing
Is Deformable Image Registration Ready for Brain Metastasis Reirradiation Dose Accumulation? A Longitudinal MRI Benchmark of Registration Accuracy
Dose accumulation is increasingly important in adaptive radiation therapy and reirradiation, but its clinical validity depends on the performance of deformable image registration (DIR). Reirradiation of brain metastases (BMs) with stereotactic radiosurgery (SRS) provides a controlled but clinically meaningful DIR test case: intra-subject brain deformation is usually limited after rigid alignment, yet recurrent lesions can undergo substantial local shape and volume changes that rigid registration cannot capture and can affect dose accumulation. We benchmarked a wide range of learning-based and optimization-based DIR methods on 87 manually screened longitudinal contrast-enhanced T1-weighted MRI lesion pairs from an institutional BM SRS retreatment cohort. Learning-based methods pretrained on healthy-brain MRI were evaluated zero-shot and after instance-specific optimization (ISO) or tumor-proximity target-specific optimization (TSO). Registration was assessed using lesion overlap (Dice), surface distance metrics (HD95 and sASD), target-volume recovery, and runtime and memory. Pretrained learning-based methods showed variable zero-shot performance, while ISO/TSO improved all tested learning-based families. However, optimization-based methods remained the best-performing approach while maintaining reasonable runtime. These findings suggest that even state-of-the-art DIR methods do not yet provide sufficiently accurate and consistent registration for unmonitored use in brain metastasis reirradiation dose accumulation. Because accurate registration is a prerequisite for deformable dose accumulation, clinical application will require case-level quality control and direct assessment of how registration uncertainty affects downstream dose metrics.
Read the full abstract
Artificial Intelligence
ORDDAR: Observation-Driven Reasoning for Distortion-Resilient Decision, Action, and Cognitive Recovery
AI agents increasingly perform long-term reasoning, planning, tool use, memory integration, and autonomous decision making, yet erroneous intermediate states can propagate and cause inconsistent decisions and unreliable outputs. Existing reasoning approaches mainly rely on iterative planning, self-reflection, augmented memory, or verification, but rarely localize and selectively repair faulty reasoning. We present ORDDAR (Observation-Driven Reasoning for Distortion-Resilient Decision, Action, and Cognitive Recovery), a reasoning framework that models reasoning as cognitive state transitions, detects localized distortions, retrieves related reasoning from prior experiences, and repairs only the affected states. ORDDAR therefore performs recovery at the local reasoning-transition level rather than regenerating the complete trajectory. Experiments across mathematical, commonsense, multi-hop, and clinical reasoning benchmarks demonstrate improved reasoning quality, recovery ability, and interpretability over multiple evaluated reasoning baselines.
Read the full abstract
Physics and Society
Save 2050: A Planetary-Scale Collective Prediction System for the Singularity Crisis
The Anthropocene mode of development is driving civilization toward a "singularity crisis": artificial intelligence (AI) is rapidly approaching general intelligence (AGI) with escalating risks of losing control, while unrestrained economic growth generates super-exponential growth of entropy production that pushes the Earth system toward its tipping points. This paper proposes the "Save 2050" initiative: a distributed planetary-scale collective prediction system that aggregates judgments about the future from humans and AI through open registration, crowdsourcing, and incentive mechanisms, and integrates them via large-scale simulation into inspectable "predicted worlds," enabling humanity to systematically see the future for the first time. We argue for the initiative's feasibility along four dimensions—the maturation of AI forecasting, human collective intelligence and the institutional environment, supporting progress in related fields, and societal demand. We then identify three key enabling technologies: long-horizon automated resolution, simulation-based prediction aggregation, and reflexivity governance. We analyze potential risks—including reflexivity, cognitive monoculture, narrative capture, and regulatory and ethical concerns—together with mitigation strategies, and we outline a phased roadmap with open problems. The initiative's primary goal is not to intervene in the future, but to make the future visible, discussable, and co-writable.
Read the full abstract
Computer Vision
Stochastic Liquid Deformation Fields: An SDE Generalisation of Closed-Form Continuous-Time Cells for Dynamic 3D Gaussian Splatting
Deformable 3D Gaussian Splatting (D-3DGS) reconstructs dynamic scenes by deforming a canonical set of 3D Gaussians through a deformation field of frame time. Replacing its MLP with a stack of Closed-form Continuous-time (CfC) cells-a Liquid Neural Network that solves the Liquid Timeconstant ODE in closed form-gives the field continuous-time behaviour at feed-forward cost. That closed form, however, is only the deterministic limit of a noise-driven system, and drops the stochastic term usually credited for the robustness of liquid networks. We put it back: a small Gaussian perturbation is added to the time gate of every CfC cell, turning the deterministic field into a simple stochastic (SDE) one. The noise is used only during training, needs no solver, and reduces exactly to the CfC when switched off. On the synthetic D-NeRF scenes the stochastic field is on par with the deterministic CfC and beats the MLP baseline on most scenes; on the real-world NeRF-DS scenes the deterministic limit is already best and adding noise does not help. The study thus gives both a clean way to read the CfC field as an SDE and an honest account of when a plain noise term helps and when it does not.
Read the full abstract
Computer Vision
TopoAgent: A Structure-Aware Perception-to-Reasoning Framework for Diagram-to-Graph Topology Extraction with Large Vision-Language Models
Diagram-to-graph topology extraction aims to extract a graph of entities and their connections from a structural diagram. This task remains challenging for current vision-language models because it requires both fine-grained perceptual grounding and topology-aware reasoning with global consistency. We present TopoBench-180, a human-verified benchmark for diagram-to-graph topology extraction, and TopoAgent, a structure-aware perception-to-reasoning framework for reliable topology extraction using large vision-language models. TopoBench-180 contains 180 structural diagrams spanning Web-style and Network-style categories, paired with canonical graph annotations. TopoAgent progressively extracts the target graph by combining grounded perception, global structural priors, canonical node inventory construction, node-centric local-to-global relation reasoning, and topological consistency enforcement. Experiments on TopoBench-180 show that TopoAgent outperforms strong vision-language model baselines and recent visual reasoning frameworks, especially on edge extraction. More broadly, this work fills an important gap in multimodal structured understanding by establishing a benchmark and framework for diagram-to-graph topology extraction. The benchmark and associated resources will be publicly released at https://huggingface.co/datasets/WayneGuo0011/TopoBench-180.
Read the full abstract
Lattice Field Theory
Machine Learning Unveils Finite-volume Energy Shifts in Three-body System
Finite-volume extrapolation (FVE) is essential for extracting physical observables in the lattice calculation. While rigorous FVE formulations are well established for short-range potentials in both two- and three-body systems, long-range interactions with force ranges comparable to the lattice size $L$ remain challenging. Extending a previous data-driven scheme for two-body systems, we apply symbolic regression (PySR) to uncover universal three-body FVE formulae. For short-range potentials, we reproduce the two limiting cases, i.e. $\kappa_3\gg\kappa_2$ and $\kappa_3\sim\kappa_2$. For pure long-range potentials, we obtain a dedicated analytic expression, and after incorporating short-range contributions, we uncover a unified formula consistent with the original PySR solution, which performs excellently in the intermediate force range around 1 fm. This work demonstrates that combining machine learning with physical constraints can yield novel analytical results inaccessible to conventional theoretical tools, advancing data-driven methodologies in hadron physics.
Read the full abstract
Computer Vision
Beyond Visual Boundaries: Rethinking Scene Segmentation for Movie RAG
Understanding long-form video remains a fundamental challenge for multimodal large language models (MLLMs). Sparse frame sampling fails to capture fine-grained visual details, while dense sampling quickly exceeds context length limits. Retrieval-augmented generation (RAG) offers a promising middle ground by selectively retrieving relevant video segments for grounded generation, yet its effectiveness critically depends on the quality of the video segments used as retrieval units. In this paper, we investigate RAG for movie understanding, which demands story-level reasoning over characters, events, and narrative arcs spanning hours of content. Scene segmentation, a long-studied problem that partitions movies into semantically coherent units, is a natural candidate for defining such retrieval units. We reexamine whether existing methods actually serve this role through comprehensive evaluation on downstream movie understanding tasks, and find that they do not provide a consistent or dominant advantage over naive uniform temporal chunking. Our audit of the most standard scene segmentation benchmarks reveals why: current annotations prioritize visually salient transitions over narrative event structure. Motivated by this mismatch, we introduce NarraScene, a narrative-centric scene segmentation dataset annotated with a three-level cognitive taxonomy spanning physical, character, and narrative change, where every valid boundary requires a narrative-level shift. When used as retrieval units, these narrative-grounded segments outperform uniform chunking on downstream movie understanding tasks, suggesting that the central challenge for scene segmentation in movie RAG is not detecting boundaries, but identifying the narrative event units that matter for movie understanding. The dataset is available at https://huggingface.co/datasets/queez0405/NarraScene.
Read the full abstract
Computer Vision
State-Conditioned Visual Evidence Retrieval for Fine-Grained Perception in Document Vision-Language Models
Compared with typical vision-language tasks, document parsing places stronger demands on fine-grained visual perception. Existing vision-language model (VLM)-based parsing approaches rely on globally compressed visual tokens, where fine-grained details are entangled within a single representation and repeatedly accessed during decoding. However, we observe that the visual evidence for each prediction is typically localized and conditioned on the current decoding state, whereas such representations must be accessed in full at every decoding step, resulting in inefficient computation. To address this mismatch, we formulate perception as state-conditioned visual evidence retrieval (SCVER) during autoregressive decoding. The model operates on a compact global representation for coarse structure and retrieves a small set of relevant high-resolution regions conditioned on the current token state. This coarse-to-fine design enables on-demand access to fine-grained visual cues, relieving globally shared representations from encoding all fine-grained details. We further find that learning such state-conditioned retrieval in VLMs is challenging and unstable. To stabilize this process, we introduce a Spatially-Guided Learning Objective (SGLO) to guide the retrieval process. Experiments on document parsing benchmarks show that SCVER improves robustness under reduced input resolution and achieves a better accuracy-efficiency trade-off, demonstrating the effectiveness of on-demand visual evidence retrieval for fine-grained perception.
Read the full abstract
Robotics
Multi-Group Pipe Routing under Permanent Geometric Occupancy: Problem, Benchmark, and Classical Baselines
Additive manufacturing (AM) enables compact hydraulic components whose internal fluid channels can follow free-form 3D paths rather than conventionally drilled holes. A representative case is multi-group channel layout in rotary direct-drive servo valves: once a channel is placed it permanently occupies volume, so later channels must clear earlier geometry–unlike classical multi-agent pathfinding (MAPF), where agents free space after moving. We study this setting as multi-group pipe routing under permanent geometric occupancy. Our contributions are a problem formalization with geometric dual-witness conflicts, a constructive 3D benchmark (two corridor generators x two obstacle painters, controlled difficulty, feasibility witnesses), and baseline results for CBS, PBS, and priority planning adapted to this coupling. We evaluate by success rate under a fixed time budget. PlaneSlice hard instances clearly rank CBS above PBS and PP, while easier cells validate the pipeline. The goal is a reproducible problem definition, suite, and classical baselines–not a new optimal MAPF algorithm.
Read the full abstract
Computer Vision
Instruction Distillation: Text Instructions as Visual Examples
Visual in-context learning (ICL) with multimodal large language models (MLLMs) is effective for fine-grained visual classification, but each retrieved image example consumes several hundred context tokens, making large-$K$ settings prohibitively expensive at inference scale. We propose Instruction Distillation: an offline procedure in which the MLLM itself generates, for each individual training image, a structured identification instruction encoding general appearance cues, features that differentiate the class from visually similar ones, and a common confusion point. Unlike prior work that produces a single description per class, our instructions are generated per training image, preserving the intra-class visual diversity that per-class descriptions collapse. At inference time, we study five configurations sharing a single CLIP retrieval index: zero-shot, image ICL, instruction-only ICL, and two hybrid variants in which retrieved neighbors are split between images and instructions. Across seven fine-grained benchmarks and two MLLM backbones, instruction based pipelines match, or exceeds image ICL at $K{=}1$ and reduces per-query tokens by $2.9\times$ and inference latency by $3.3\times$ at $K{=}5$. Hybrid configurations further show that visual and textual ICL signals are complementary, images give visual patterns to learn and see, while instructions give explicit rules and logic. When both of these are provided, the quality of context improves, which is noticeable in the performance.
Read the full abstract
Computer Vision
Weaving Visual Narratives: Agentic Image Bundle Composition Beyond Atomic Visual Matching
Image retrieval has traditionally been formulated as a point-wise matching problem, where each candidate image is scored in isolation. However, this atomic paradigm fails to capture the complexity of human search intent within personal photo collections, where users often seek compact visual stories bound by structural relations rather than isolated snapshots. To address this limitation, we introduce **Image Bundle Composition (IBC)**, a novel paradigm that shifts the objective from ranking individual images to dynamically composing cohesive image bundles from a massive, unstructured photo pool. Since target bundles are not predefined, IBC presents a severe combinatorial explosion challenge and demands modeling non-decomposable joint relevance. To establish this paradigm, we construct **IBCBench**, the first IBC benchmark dataset containing 109,467 images and 667 verified queries, built via a semi-automated verification pipeline. Furthermore, we propose **BundleWeaver**, an agentic framework that reformulates IBC as query-conditioned incremental hyperedge discovery. By employing a Large Language Model to adaptively search for missing relational roles and utilizing a Vision-Language Model for whole-bundle verification, BundleWeaver effectively navigates the combinatorial space. Extensive experiments demonstrate that while state-of-the-art embedding models and static decompose-and-rerank paradigms suffer from relational blindness, BundleWeaver achieves substantial performance gains, highlighting the necessity of shifting from atomic scoring to dynamic relational composition. Our dataset and code are available.
Read the full abstract
Computer Vision
SNF-Bench: Separating Static Drift from Natural Flow in Long-Horizon Fixed-Camera Video Generation
Long-horizon video generation is evaluated with whole-frame metrics that reward motion and temporal consistency. For fixed-camera nature scenes this creates an ambiguity: motion of water, fire, smoke, or rain is desirable, whereas motion of the background is an error. A system can therefore score well on motion while its scene drifts, or on consistency while its flow stagnates. We introduce SNF-Bench, an evaluation framework for long-horizon fixed-camera generation that partitions each scene into static support and dynamic flow and reports static fidelity, flow persistence with absolute magnitude, and drift leakage separately, never as one score. Drift leakage is interpretive context rather than a headline measurement. Each factor is validated mechanistically rather than by correlation with preference: we inject global translation, rotation, and scale drift and progressive late freezing at known severity into real generations, and require each factor to respond in its stated direction and to remain selective against corruptions it does not target. Auditing publicly released long-horizon text-conditioned checkpoints under one recorded common inference configuration, plus an image-conditioned track with released-pipeline references and a deployment-sensitivity panel, we find that whole-frame motion and static-region drift induce near-opposite orderings of the same outputs. At maximum controlled translation, fBD and NBF rise to $1.86\times$ and $1.32\times$ baseline, but whole-frame Dynamic Degree reaches only $1.07\times$—rewarding the corruption. SNF-Bench measures where motion occurs and whether it persists; it does not measure physical realism. Project page: https://minar09.github.io/snfbench/.
Read the full abstract
Robotics
RoboGesture: Real-Time Semantic-aligned Co-Speech Gestures Generation for Humanoid Interaction
Enabling humanoid robots to respond to human speech with synchronized and semantically meaningful gestures is fundamental to natural human-robot interaction. However, this task faces three critical barriers: the scarcity of semantically rich datasets, the "modality eclipse" where models ignore audio cues in favor of kinematic inertia, and the sim-to-real gap regarding physical safety. We propose RoboGesture, a robot-centric framework that co-designs data, modeling, and control to power a complete interactive human-humanoid system in which the robot listens, responds, and gestures in real time. We first establish the RoboGesture dataset featuring over 300 gesture categories and develop an automated pipeline to synthesize large-scale collision-free, robot-specific audio-motion pairs. Our architecture features a Hierarchical Semantic-Acoustic Aligner that extracts multi-granular prosodic and semantic cues directly from raw audio tokens. These cues drive a Streaming Conditional Motion Generator based on a diffusion transformer with conditional flow matching. To ensure high responsiveness, we introduce Anti-Inertia CFG Masking, which prevents the model from collapsing into repetitive historical patterns by compelling it to proactively mine control signals from the audio modality. Finally, an MPC-based safety filter ensures real-time, collision-free execution on physical hardware. Experiments on a Unitree G1 humanoid demonstrate that RoboGesture generates safer, more rhythmic, and more semantically appropriate responses compared to state-of-the-art baselines.
Read the full abstract
Computer Vision
Growing a Stand, Not a Tree: Joint Canopy Generation Reproduces Crown Shyness
In closed forests, neighboring tree crowns often stop short of touching, leaving a network of narrow gaps known as crown shyness. The pattern belongs to the stand rather than to any single tree, which makes it a natural probe of a question in generative modeling: can a learned model produce a set of objects whose defining structure exists only between them? We formulate stand-level canopy generation as set generation with a flow-matching model, in which attention between trees is the only channel through which coupling can arise. Trained on stands grown by a resource-competition simulation that is provably not reducible to per-tree geometry, the joint model halves the clearance distribution error of an identical-capacity model that generates each tree alone, and the advantage persists at stem densities outside the training range. Against field measurements of a tropical oak forest, a single calibrated scalar yields held-out agreement in gap magnitude and crown asymmetry. The directional statistics of the gaps are controlled by stem placement rather than by the growth rule, and match the field once stem jitter is calibrated. Crown shyness, in both the simulation and the learned model, is a property of the stand and not of the tree.
Read the full abstract
Quantum Physics
Scaling Alternating-Bias-Assisted Annealing for Precision Transmon Frequency Targeting on Superconducting Quantum Processors
Recent advances in the alternating-bias-assisted annealing (ABAA) technique have successfully mitigated intrinsic Josephson-junction (JJ) fabrication variations. This new technique enables precision qubit frequency tuning alongside simplicity. However, it is critical to enhance tuning throughput and yield while investigating the factors that drive targeting performance as the technology scales. Here, we characterize ABAA tuning performance within a 150-mm wafer process flow and extend this technique to simultaneous, multi-channel tuning, demonstrating that a wafer-scale JJ resistance tuning precision of $\sigma=0.50\pm0.05\%$ alongside a component-level yield of $\ge 98.8\%$ can be achieved. Furthermore, we demonstrate a strong correlation between yield, tuning speed, and junction breakdown voltage, establishing the latter as a vital process control parameter for meeting production goals. Finally, we demonstrate a successful implementation of ABAA tuning on a quad-module quantum processor (Rigetti Cepheus-1-36Q), where we achieve an empirical frequency targeting precision of $\sigma \sim 30\text{ MHz}$ in both qubit and qubit-qubit detuning frequencies, contributing to high median two-qubit gate fidelities. These results confirm the efficacy and scalability of ABAA for high-precision Hamiltonian targeting, a critical enabler for modular superconducting quantum processor technology.
Read the full abstract
Combinatorics
Default-Distance Entropy and Metric Dimension in Finite Geometries
A resolving set in a graph is a set of landmarks whose distance vectors distinguish all vertices. We use information theory to prove lower bounds for metric dimension and class dimension in distance-regular graphs and association schemes arising from finite geometry. The core idea is that, for a fixed landmark, a random object usually lies in one overwhelmingly likely distance or relation class. For classical dual polar graphs, with rank and type fixed and $q\to\infty$ through the admissible field orders, we prove $\mu(\Gamma(q,d,e))=\Theta_{d,e}(q^e)$ for $d\geq 2$ and $e>0$. The lower bound uses opposition as the typical distance. For the upper bound, we take, for each of a constant number of $(d-1)$-dimensional singular subspaces, all generators containing it. For Grassmann graphs, bilinear forms graphs, and attenuated-space schemes, we obtain lower bounds of the same exponential order as the known incidence constructions.
Read the full abstract
High Energy Astrophysical Phenomena
Low-energy antinuclei measurements for background-free indirect dark matter searches and PBH signatures
Light low-energy cosmic ray antinuclei constitute powerful probes for detecting dark matter or other sources of new physics in our Galaxy. This ASTRA Initiative Mission Concept reviews the theoretical motivation, highlighting the transformative potential and current experimental status. Motivated by exciting tentative results, it makes a clear case for the need for the next flagship mission to confirm or refute claims of cosmic antinuclei as signs of new physics.
Read the full abstract
Instrumentation and Methods for Astrophysics
JUG: JAX-based Unified pulsar timinG
We present JUG (JAX-based Unified pulsar timinG), a JAX-based, fully independent pulsar timing package emphasising speed and ease of use, designed to confidently handle the increasingly large and complex pulsar timing array datasets that are being created in the pulsar timing field. JUG implements the entire pulsar timing pipeline itself, from data handling and clock corrections through to the timing model and fitting, without relying on other timing software. It enables Pythonic programming at the speed of compiled code, is GPU-capable, and can be operated via a Python API or an interactive GUI. A user of JUG can interactively explore data, fit timing models with complex stochastic noise, model deterministic signals such as continuous gravitational waves, and obtain accurate point estimates of the parameters of the stochastic processes present, thereby bridging frequentist timing and Bayesian noise analysis. JUG is faster than PINT by more than fifty times and is comparably fast to Tempo2, can handle millions of arrival times, agrees with PINT at the picosecond level, and can reliably recover known timing model and noise parameter values. In this paper we describe its design, performance, and validation, and demonstrate its advantages for pulsar timing data analysis.