Peer Revenue Fully funded

Papers

Fresh from arXiv, typeset properly. Highlighted words trade on the market.

All of arXiv

Sunday 27 September

Computer Vision

PARSEE-VAD: Efficient Training-Free Online Video Anomaly Detection via Proposition-Aware Reasoning and Streaming Evidence Escalation

Ji Wang, Shuangqing Zhang, Guo-Sen Xie et al. (+1 more)

Training-free online video anomaly detection (VAD) with frozen multimodal language models faces two coupled challenges: extracting reliable current-window semantics under causal and computational constraints, and maintaining temporal continuity without repeatedly transmitting high-dimensional history. Encoding history through text can compress visual evidence and introduce semantic bias, whereas retaining visual history expands multimodal context. We introduce PARSEE-VAD, a two-module framework that separates semantic evidence acquisition from score-state evolution. Proposition-Aware Reasoning (PAR) extracts structured propositional evidence from the current causal window and conditionally activates more specific queries when coarse evidence warrants further refinement. By sharing a reusable causal visual prefix across queries, PAR reduces redundant computation through selective execution. Streaming Evidence Escalation (SEE) maps the acquired proposition evidence into a compact score-domain event state through current evidence escalation, then propagates only the resulting bounded state across decisions to support temporal continuity. Experiments on four benchmarks demonstrate strong training-free online performance while selective routing reduces specialist computation and score-state propagation remains sparse. These results support a current-first principle for streaming multimodal inference: resolve present semantics first, then use compact historical state only to repair residual continuity gaps.
Read the full abstract

All authors Ji Wang, Shuangqing Zhang, Guo-Sen Xie, Fang Zhao

Robotics

Learning with Object-centric Representations of Tactile Interactive Perception for Robot Manipulation

Xinyi Yang, Zilin Si, Zhuowei Xu et al. (+2 more)

Implicit object properties that are difficult to directly infer from vision, such as material, container contents, or softness, can be revealed through tactile sensing and exploratory interactions. However, because tactile signals are transient and sparse, extracting informative tactile events and effectively incorporating them into robotic manipulation remains a challenge. In this work, we present an object-centric context-aware manipulation framework that learns task-agnostic object representations through tactile exploration. A token learner autonomously selects representative tactile segments from long-horizon exploration, while contrastive alignment with descriptive text embeddings enables a latent space that captures multiple physical object properties. These learned representations are then used as semantic context to guide object-centric manipulation policies and adapt strategies based on object properties. Experiments show that the learned representations achieve 93% and 84% property estimation accuracy on seen and unseen objects. Evaluated on three tasks involving visually ambiguous objects, i.e. multi-object rearrangement, pouring, and box opening, the proposed framework improves both target selection and property-dependent manipulation adaptation, raising task success, aggregated over all evaluation trials, from 41% to 92% on seen objects and from 19% to 67% on unseen objects over baseline policies without object-context conditioning. Videos and additional results are available at https://xinyiyxyx.github.io/tactile-object-centric/.
Read the full abstract

All authors Xinyi Yang, Zilin Si, Zhuowei Xu, Zeynep Temel, Oliver Kroemer

Game Theory

Fair Share Allocations for Almost All Agents

Tomer Ezra and Tamar Garbuz

We study fair allocation of indivisible goods among agents with additive valuations. Our main result guarantees every agent her exact AnyPrice Share (APS) whenever the total entitlement is at most $1-\varepsilon$ and each individual entitlement is sufficiently small. For equal entitlements, this yields a $1$-out-of-$(n+O(\sqrt{n\log n}))$ maximin share (MMS) allocation. We also establish two best-of-both-worlds guarantees that preserve proportionality in expectation: each agent receives her full MMS with probability at least $1-O((\log n/n)^{1/3})$, or every agent receives her $1$-out-of-$(n+O(n^{2/3}(\log n)^{1/3}))$ MMS in every outcome. Both guarantees extend to sufficiently small unequal entitlements, with APS replacing MMS. Quantitatively, our deterministic result allows an entitlement cap of order $\varepsilon^2/\log(1/\varepsilon)$, while an impossibility construction shows that any universally sufficient cap must be $O(\varepsilon)$.
Read the full abstract

Information Retrieval

Inspire: Benchmarking Scientific Literature Search for Open Research Problems

Jianrong Ding, Zhengyan Shi, Jianyuan Zhong et al. (+5 more)

Scientific literature search often begins with an open research problem rather than a known target paper or a fixed candidate set. We introduce INSPIRE, a benchmark for evaluating agents that search prior literature to make progress on solution-redacted research problems. Each instance pairs a research brief with a target-specific cutoff three months before a later paper and evaluates ranked outputs against graded cited antecedents from that paper's realized research lineage. Search proceeds over an open corpus, while the identity of the target paper and membership of its cited antecedents remain hidden from the agent. Beyond end-to-end retrieval quality, INSPIRE uses logged search trajectories to distinguish three coupled stages: resource exposure, whether useful antecedents are surfaced during search; selection, whether exposed antecedents are retained; and ranking, how effectively retained papers are ordered. Across 476 computer-science targets under a shared search interface and budget, the strongest evaluated agent achieves 0.284 nDCG@10. Results show that current agents more readily recover an isolated antecedent than assemble a broader portfolio of relevant prior work. The stagewise analysis identifies resource exposure as the largest observed bottleneck, with further losses in selection and ranking. We additionally construct replay-valid hindsight demonstrations and show that they improve held-out search without changing test-time information, establishing that the benchmark provides an actionable learning signal. INSPIRE therefore enables both end-to-end comparison and stage-resolved diagnosis in a setting where the agent must construct its own working criterion of relevance.
Read the full abstract

All authors Jianrong Ding, Zhengyan Shi, Jianyuan Zhong, Kai Qiu, Qi Dai, Yifan Yang, Chong Luo, Qiang Xu

Machine Learning

MTLiquid: Enabling Efficient Multi-Task Learning using Liquid Neural Networks for Lightweight Healthcare Monitoring Systems

Rachmad Vidya Wicaksana Putra, Fahad Abdul Rauf and Muhammad Shafique

Continuous-time sensing and monitoring with timely and accurate decision-making are critical for many real-world applications. In healthcare monitoring systems, physiological signals are often available or sampled at irregular time intervals, hence requiring continuous-time processing to provide accurate prediction. Moreover, such systems often need to solve multiple detection/prediction tasks to provide a comprehensive patient review from different physiological aspects for more accurate decision-making. To solve this, continuous-time neural networks (CTNNs) can be employed. However, state-of-the-art works typically solve only one task at each network, thereby limiting their efficiency gains. To address this limitation, we propose MTLiquid, a novel methodology to enable efficient multi-task learning in continuous-time processing for healthcare monitoring systems through effective network design and training strategy. MTLiquid employs: (1) multiple input and output heads to accommodate different tasks, while sharing the same backbone network across tasks; as well as (2) an effective training strategy that leverages a loss-weighting technique to balance learning updates across different tasks and a proportional data presentation technique to address imbalanced dataset sizes. Experimental results for mortality prediction (P12) and sepsis early detection (P19) tasks for ICU patients show that, MTLiquid achieves strong performance (AUROC: 0.84 for P12 and 0.94 for P19) comparable to the state-of-the-art single-task learning in both continuous-time networks (AUROC: 0.84 for P12 and 0.95 for P19) and discrete-time networks (AUROC: 0.79-0.82 for P12 and 0.92-0.94 for P19), while incurring significantly smaller memory cost by 44%-94%. These results highlight the potential of our MTLiquid methodology to enable lightweight continuous-time healthcare monitoring systems for better decision-making.
Read the full abstract

Quantum Algebra

Inequivalent modular tensor categories with identical $S$, $T$ and $W$

Ran Luo and Jiahua Tian

It remains an open question whether the modular data $(S,T)$ together with the Whitehead-link matrix $W$ determine a modular tensor category up to ribbon equivalence. We answer negatively by constructing two braided-inequivalent Dijkgraaf–Witten modular categories for $(\mathbb Z/p\mathbb Z)^3$ for prime $p\ge5$ with identical $(S,T,W)$ and, more generally, identical Reshetikhin–Turaev invariants of all framed oriented links with at most two components under a single bijection of simple objects. The construction varies an alternating $3$-cocycle while fixing a quadratic cohomology class, using restriction to subgroups of rank at most two to match link invariants and the quadratic and alternating classes together to obstruct braided equivalence. A complete classification of modular tensor categories up to ribbon equivalence therefore requires invariants beyond those of colored framed links with at most two components. We also analytically evaluate an uncolored partition function on a closed oriented three-manifold that separates all $(p-1)/2$ equivalence classes in the family.
Read the full abstract

Computer Vision

AevaScenes: An FMCW LiDAR Dataset and Benchmark for Long-Range Perception

Gautham Narayan Narasimhan, Heethesh Vhavle, Kumar Bhargav Viswanatha et al. (+2 more)

FMCW LiDAR measures per-point radial Doppler velocity alongside range, providing a motion cue unavailable in conventional time-of-flight sensors. Exploiting this signal at long range remains understudied. We present an FMCW LiDAR dataset of 575 sequences (57.5K frames) with over 8 million annotated 3D boxes across 16 detection classes and per-point labels across 24 semantic classes, captured by six commercial FMCW LiDAR sensors and six paired 4K cameras across eight Bay Area cities, including 237 nighttime sequences, with annotations extending to 400m. We define a benchmark with three tasks: 3D object detection, scene flow estimation, and semantic segmentation. Detection and scene flow are evaluated across three range bins to 400m, with a public evaluation server. We explore the impact of Doppler measurements on flagship recognition tasks, and find significant improvements up to 2X in detection AP of far-away vehicles and pedestrians, particularly in low-latency single-frame settings. We similarly find scene flow accuracy is significantly improved with Doppler measurements across all ranges. Our dataset and benchmark have been publicly released at https://scenes.aeva.com.
Read the full abstract

All authors Gautham Narayan Narasimhan, Heethesh Vhavle, Kumar Bhargav Viswanatha, James Reuther, Deva Ramanan

Quantum Physics

In-plane optimal qubit operation with control fidelities exceeding 99 %

Yu-Chen Zhou, Zhenzhen Kong, Rong-Long Ma et al. (+9 more)

Hole spin qubits based on semiconductor quantum dots are promising for building future large-scale quantum computers owing to their all-electrical manipulation. However, abundant physical mechanisms of valence band holes lead to anisotropic qubit properties. There is an opportunity to prolong the coherence time and achieve high-fidelity qubit manipulations. Here, we report a single-hole spin qubit in a planar germanium quantum dot and investigate its anisotropic susceptibility to charge noise under an in-plane magnetic field. By correlating the longitudinal spin-electric susceptibility with qubit coherence and control performance, we identify an optimal operating point where the sensitivity to charge noise is minimized. We find that optimizing the magnetic-field orientation reduces the spin-electric susceptibility, resulting in a five-fold enhancement of the Hahn-echo coherence time and a nearly tenfold suppression of control infidelity. At the optimal operating point, gate set tomography demonstrates the maximum gate fidelity of 99.82 %. Our finding enhances the prospects of hole spin qubits for scalable quantum information processing.
Read the full abstract

All authors Yu-Chen Zhou, Zhenzhen Kong, Rong-Long Ma, Hao-Tian Jiang, Ao-Ran Li, Yang Zhong, Zhi-Tao Wu, Rui Zheng, Gui-Lei Wang, Gang Cao, Hai-Ou Li, Guo-Ping Guo

Quantum Physics

Occupation-selective photon-pair interactions in a flux-pumped SQUID resonator

L. N. Ferreira, D. Z. Rossatto, A. S. M. de Castro et al. (+1 more)

Flux-pumped Josephson circuits provide a versatile platform for parametric photon generation and multimode bosonic control. We identify an occupation-conditioned pair interaction that arises when two harmonic normal modes participate in the same driven SQUID coordinate. We consider a degenerate pair process in which two photons are created or annihilated in a signal mode while the occupation of a second harmonic mode remains unchanged. Although it does not exchange excitations, this controller mode modifies both the resonance frequency and the strength of the pair transition. The two effects have different physical origins: diagonal Josephson nonlinearities shift the pair resonance, whereas the nonlinear Josephson matrix element makes the avoided-crossing gap itself depend on the controller occupation. Starting from the full Josephson cosine, we develop complementary time-dependent Schrieffer–Wolff and Magnus descriptions that expose the direct and virtual processes responsible for this interaction, and benchmark the resulting effective theory against a numerically converged Floquet treatment of the unexpanded two-mode circuit Hamiltonian. The occupation dependence produces a two-dimensional pair spectrum that enables selective transitions within the bosonic Fock lattice. When the controller is prepared in a coherent superposition, the same interaction converts this conditional dynamics into signal–controller entanglement and controller-dependent non-Gaussian state generation. Observable-level comparisons further show that the variation of the pair-transition strength cannot be reproduced by occupation-dependent resonance shifts alone. These results establish a boson-controlled parametric interaction in which the state of one harmonic mode controls both the spectral position and the coupling strength of a nonlinear process in another.
Read the full abstract

All authors L. N. Ferreira, D. Z. Rossatto, A. S. M. de Castro, G. D. de Moraes Neto

Signal Processing

Index Modulation Aided Phase-Rotated AFDM for Efficient Integrated Sensing and Communication: Framework, Optimization and Performance Analysis

Shiqi Cui, Zeping Sui, Tianqi Mao et al. (+7 more)

Affine frequency division multiplexing (AFDM) has emerged as a competitive waveform for integrated sensing and communication (ISAC) in high-mobility scenarios, but its data-dependent high peak-to-average power ratio (PAPR) and ambiguity function (AF) sidelobes may degrade the power-amplifier efficiency as well as the sensing performance. To mitigate the issues, one promising candidate is the phase-rotation optimization philosophy, which necessitates auxiliary side information (SI) to guarantee successful data recovery at the receiver. However, using additional signalling overhead to deliver the SI degrades the spectral efficiency. Against the background, this paper proposes a power level index modulation (PLIM)-aided phase-rotated AFDM (PLIM-PR-AFDM) scheme, which embeds the phase-rotation SI within the transmitted block. At the transmitter, the proposed scheme jointly optimizes the waveform by coordinating the PLIM power allocation, phase rotation, and embedding of SI. Specifically, subcarriers are flexibly allocated with different transmit power levels. High-power subcarriers are phase-rotated to suppress the PAPR and fast-slow-time AF sidelobes. By decomposing the fast-slow-time AF sidelobe energy into per-block periodic autocorrelation function (PACF) terms, the sensing objective can be formulated as a PACF-weighted integrated sidelobe level (WISL) minimization problem, while selected low-power subcarriers embed the phase bits as SI through constellation labels. Furthermore, the PLIM power patterns of each transmitted block can convey additional index bits, which compensates for the spectral efficiency loss caused by SI transmission. Moreover, the error performance is analyzed by deriving an upper bound of the average bit error probability (ABEP) and identifying the condition for full diversity.
Read the full abstract

All authors Shiqi Cui, Zeping Sui, Tianqi Mao, Yuanshuo Gang, Guangyao Liu, Fan Zhang, Miaowen Wen, Dezhi Zheng, Christos Masouros, Zhaocheng Wang

Computation and Language

Beyond Memory Construction: Rethinking Memory Access for LLM-based Conversational Agents

Donghua Cai, Yongheng Deng, Yifei Wang et al. (+2 more)

Memory is a core component of conversational agents, enabling coherent and context-aware behavior over long interactions. Recent approaches commonly rely on LLM-based memory construction, where raw interactions are rewritten into structured memory units and later retrieved via a RAG pipeline. While effective in controlled settings, we show that this paradigm breaks down in long-horizon, high-entropy conversations: memory construction becomes increasingly lossy and unstable as context length and information complexity grow, and incurs prohibitive cost due to repeated LLM invocation. To address these limitations, we propose Threader, a memory system that shifts the focus from memory construction to efficient, structure-aware access over raw interactions. Instead of rewriting interactions, Threader preserves them as first-class memory, organizes them into topic-coherent segments via lightweight incremental segmentation, and enables accurate retrieval through multi-view representation. At query time, it performs multi-signal retrieval that combines segment-level access with localized evidence matching, ensuring both completeness and coherence. Extensive experiments demonstrate that Threader consistently improves answer accuracy and evidence recall, while significantly reducing the memory construction overhead.
Read the full abstract

All authors Donghua Cai, Yongheng Deng, Yifei Wang, Zijun Shen, Ju Ren

Distributed Computing

PackServe: SLO-Aware Request Scheduling for Agentic LLM Serving at Scale

Zhiyuan Tan, Dejiang Zhu, Jingzhe Jiang et al. (+4 more)

Request scheduling is a key challenge in large-scale clusters serving agentic large language model (LLM) workloads. An effective scheduler must preserve key-value cache (KVC) reuse across long, shared prefixes, meet token-level latency service-level objectives (SLOs), and minimize GPU resource footprint. Existing schedulers struggle to reconcile these requirements: request consolidation can sacrifice cache locality and increase prefill/decode interference, compromising both SLO attainment and resource efficiency. We present PackServe, a scheduler designed to reduce resource costs while meeting latency SLOs for agentic LLM serving. PackServe uses compact white-box models to predict latency under prefill/decode interference. Guided by these predictions, it packs requests onto fewer serving instances while preserving KVC reuse and SLO constraints, trading available latency headroom for improved per-instance throughput. Evaluation on 64 NVIDIA H20 GPUs shows that PackServe uses up to 16.8% and 24.6% fewer GPU-hours than state-of-the-art schedulers under 30-ms and 50-ms TPOT targets, respectively, while meeting the target TPOT objectives. PackServe has also been deployed in our production cluster comprising over 1000 GPUs, where it reduces the resource footprint by 34.7% compared with the original production scheduler.
Read the full abstract

All authors Zhiyuan Tan, Dejiang Zhu, Jingzhe Jiang, Yihao Zheng, Yang Tian, Tao Wang, Minchen Yu

Number Theory

The Beilinson–Bloch heights of generalized Ceresa cycles

Kaiyuan Gu

We compute the Beilinson–Bloch height of the generalized Ceresa cycle $\mathrm{Ce}(V)=\frac 1 2(V-[-1]_*V)$ for any cycle $V$ of pure dimension $d$ on an abelian variety $A$, in terms of an arithmetic intersection number of a certain adelic line bundle. As an application, we prove that this height is given by an adelic line bundle over algebraic families.
Read the full abstract

Astrophysics of Galaxies

Barred Galaxies in MaNGA: Stellar Mass Dependence and Black Hole Growth

Chandan Watts and Sudhanshu Barway

This study investigates the role of internal structures, such as stellar bars and active galactic nuclei (AGN), in shaping galaxy evolution using 7,408 galaxies (2,641 barred) from the SDSS-IV MaNGA survey. Our analysis spans a wide stellar mass range, from dwarf to massive galaxies, examining how internal and environmental processes regulate galaxy evolution. Galaxies are categorized by morphology (ellipticals (E), lenticulars (S0s), early-type spirals (ETS), and late-type spirals (LTS)), environment, and the presence of a bar. We find that barred galaxies have higher median specific star formation rate (sSFR) than unbarred galaxies across most stellar mass intervals. This difference varies with host morphology, with LTS and S0 galaxies showing different trends across stellar mass intervals. The environmental trends of barred and unbarred galaxies are broadly similar within the same mass intervals. In AGN hosts, barred Seyferts tend to lie near the star-forming$-$green valley boundary, while barred LINERs preferentially occupy the green valley. We further examine the relation between stellar bars and black hole mass and find that barred galaxies host systematically lower black hole masses than unbarred galaxies, with environmental effects appearing to play a secondary role in shaping these differences. A multivariate regression analysis also shows that the positive bar$-$sSFR association remains statistically significant for star-forming galaxies after accounting for stellar mass, local density, AGN activity, and morphology. Our results highlight the interconnected roles of stellar bars, stellar mass, morphology, and environment in shaping galaxy star formation, providing new insights into the role of stellar bars in galaxy evolution.
Read the full abstract

Machine Learning

RMB: Reward Model Boosting Mitigates Reward Hacking

Jiabin Fan, Dezhi Ye, Yongchang Hao et al. (+1 more)

Reinforcement Learning from Human Feedback (RLHF) is a powerful technique for aligning large language models (LLMs) with human preference. However, it often suffers from the reward hacking issue, where policy optimization improves the proxy reward model while actually degrading performance with respect to the true human preference, due to the imperfection of the proxy. To address this, we propose Reward Model Boosting (RMB), a novel approach that enhances the robustness and reliability of the reward signal for RLHF. RMB first trains a set of reward models with a diversity-promoting regularizer. This encourages each model to learn complementary aspects of the reward landscape. Then, RMB learns a lightweight aggregator in the principle of boosting to aggregate the outputs of the diverse reward models into a more accurate and robust reward signal. Our extensive experiments demonstrate that RMB significantly improves reward accuracy on both in-distribution and out-of-distribution datasets, substantially mitigating the reward hacking issue and ultimately improving RLHF performance.
Read the full abstract

All authors Jiabin Fan, Dezhi Ye, Yongchang Hao, Lili Mou

Machine Learning

When Do Models Admit They Are Wrong? Failure Disclosure Is Unstable Under Reinforcement Learning

Steven Y. Feng, Noah D. Goodman, Michael C. Frank et al. (+3 more)

Outcome-based reinforcement learning can produce models with similar task performance but very different ways of communicating about their mistakes. We study failure disclosure: whether a model admits that an attempted solution failed rather than staying silent or presenting it as successful. Across repeated outcome-only GRPO training runs, failure disclosure varies far more than task accuracy. The pattern extends to a second reasoning task and stabilized PPO, persists at 7B, and also appears in an instruction-conditioned 32B setting. We also find that small floating-point and sampling differences during training can redirect reporting behavior even when the task objective and earlier training history are held fixed. Additional tests show that failure disclosure is not a single decision: Checking the answer, entering a report, and completing the admission can separate, and the weak point depends on the task and response format. Further, experiments with neutral controls show more broadly that behaviors left weakly constrained by training are especially likely to vary across runs, of which failure disclosure is an example. We can reduce variability in failure disclosure by discouraging the model from drifting from its starting policy on failed, well-formed responses. This makes reporting substantially more consistent, though its effect on task performance depends on the setting. Stable task accuracy therefore does not guarantee stable safety-relevant behavior: Researchers should measure these behaviors directly across runs and design training methods that keep them reliable.
Read the full abstract

All authors Steven Y. Feng, Noah D. Goodman, Michael C. Frank, Evan Hubinger, Paul C. Bogdan, Andrew Lampinen

Optimization and Control

Tensor nuclear norm minimization with linear constraints

Renjie Chen, Nanxi Zhang and Bo Jiang

We study the tensor nuclear norm minimization with general linear constraints, a fundamental model for low-rank tensor optimization that includes many practical applications in computer vision, recommendation systems, revenue management, transportation research. Although the tensor nuclear norm is a convex surrogate for the tensor rank, optimizing it remains computationally challenging because evaluating the nuclear norm itself is NP-hard. We propose a semidefinite programming framework that exposes a finite structure hidden in this problem, even though it appears to possess infinitely many constraints. In particular, we show that the tensor nuclear norm admits an exact SDP representation supported by a finite number of points on the unit sphere. Based on this representation, we develop two complementary solution approaches. The first replaces the unknown support set by a finite sphere discretization which yields a solution with a provable worst-case approximation guarantee. The second selects spherical points adaptively, returns an exact optimum upon finite termination, and otherwise converges asymptotically to an optimal solution. Numerical experiments on synthetic tensor completion instances and real highway traffic data demonstrate that the adaptive method achieves competitive, and often substantially improved, recovery accuracy relative to existing methods.
Read the full abstract

Computer Vision

Scope-WM: Scoped Computation for Efficient Visual World Models

Chunzheng Li, Zesheng Jia, Hongda Zhang et al. (+4 more)

Visual world models enable robotic planning by predicting future observations, but dense latent-state propagation and sample-intensive trajectory optimization incur high inference latency and peak memory usage, limiting real-time deployment on resource-constrained platforms. Existing sparse world-model acceleration methods either rely on unguided token sparsification, which may discard planning-relevant information and restrict achievable sparsity, or introduce heavy auxiliary modules and cumbersome multi-stage training pipelines. In this work, we present Scope-WM, an efficient visual world model that scopes computation to prediction-relevant latent regions and promising action sequences. Scope-WM distills prediction relevance into a lightweight action-conditioned selector and applies full dynamics prediction only to a compact subset of selected tokens. It updates the remaining tokens using a compact summary of foreground states and their changes, allowing the background to perceive foreground dynamics without costly token-to-token interactions. During planning, Scope-WM preserves and reuses high-quality action sequences discovered during the initial MPC search, focusing subsequent search under reduced rollout budgets. The resulting pipeline requires only a one-off selector distillation followed by a single joint training stage for the sparse world model. On the challenging Push-T task, Scope-WM reduces peak GPU memory usage and planning time to $18.1\%$ and $14.3\%$ of those of dense DINO-WM, respectively, corresponding to a $6.97\times$ planning speedup, while maintaining competitive task performance. Further evaluations across five diverse visual planning tasks demonstrate the general applicability of Scope-WM. Code is available at https://github.com/ChunZheng2022/Scope-WM.
Read the full abstract

All authors Chunzheng Li, Zesheng Jia, Hongda Zhang, Jiaying Tang, Yuntian Wang, Siao Liu, Jin Wang

Computer Vision

RepFlow: Reciprocal Supervision Improves Generation and Representation in Flow Models

Weili Zeng, Feng Tian, Shengqi Liu et al. (+1 more)

Generative models learn visual structure through denoising, yet their internal states are entangled with both noise level and network depth, making it difficult to obtain a stable visual representation from the generator itself. We introduce RepFlow, which learns such a representation from the generator's evolving computation and uses it to guide generation. Specifically, a separate timestep-free encoder is trained, through a timestep-conditioned predictor, to recover generator states across depths and noise levels from masked clean images. By excluding the reference noise and masked-out content from the encoder's input, we encourage the encoder to distill visual information that is recoverable from the visible image context and predictive of generator states. The representation learned from the generator's evolving states is then fed back to guide and improve the generator, whose updated states provide supervision for further representation learning, forming a reciprocal learning process. This reciprocal process improves multi-step generation and representation quality, as measured by frozen linear probing on ImageNet with latent-space SiT and pixel-space JiT, without an externally pretrained representation teacher. Across the two unconditional settings, FID decreases by $19.2$–$40.8\%$ relative to native training, while linear-probe accuracy improves by 6.80–10.04 percentage points over the best searched raw generator features. The learned representation also serves as a distributional metric for one-step JiT post-training, extending its role from instance-level alignment to distribution-level supervision.
Read the full abstract

All authors Weili Zeng, Feng Tian, Shengqi Liu, Yichao Yan

Numerical Analysis

Hierarchical low-rank compression exploiting symmetric boundary integral equations in transmission scattering

Yasuhiro Matsumoto, Taizo Maruyama, Qianxiang Ma et al. (+1 more)

We propose an approach using matrix symmetry to accelerate hierarchical low-rank compression for wave transmission scattering. For Helmholtz transmission problems, we introduce a symmetric Müller-like boundary integral equation formulation. Although not of the second kind, this formulation creates no difficulties for direct solvers. A one-point corrected quadrature in a Nyström discretization preserves the complex symmetry of the discretized system. This formulation is integrated into a variant of a hierarchical semiseparable direct solver. Within this framework, the computational workload of the low-rank approximation in the proxy method is halved by exploiting the symmetric relationship between the left and right coefficients. Furthermore, we extend this symmetry-exploiting approach to elastic transmission problems by using a symmetric Galerkin boundary element method. Numerical experiments on a multi-node computing cluster demonstrate that the symmetric Müller-like formulation for Helmholtz scattering achieves on average a 1.41-fold speedup compared to the non-symmetric baseline while maintaining third-order convergence. Furthermore, the symmetry-aware solver for elastic scattering is more than 1.5 times faster.
Read the full abstract

All authors Yasuhiro Matsumoto, Taizo Maruyama, Qianxiang Ma, Rio Yokota

Theoretical Economics

Learning from a Mixture of Information Sources

Nicole Immorlica, Brendan Lucier, Yaroslav Mukhin et al. (+2 more)

We often learn from multiple sources that convey information in different ways. How informative is it to know the source of a signal, and how is this informativeness shaped by the distribution of sources? We extend the standard (binary-state, binary-signal) Blackwell experiment model by introducing a commonly known distribution over signaling schemes, representing the distribution of information sources. We compare learning under two information models: source-aware, where decision makers observe a signaling scheme and its realization (e.g., raw reviews, search results), and source-blind, where only the signal realization is observed (e.g., aggregate ratings, LLM-generated summaries). We show that a mean-preserving spread in the distribution of signaling schemes translates into Blackwell dominance for source-aware decision makers, implying they are "risk-loving" in information sources. In contrast, it has no impact on source-blind decision makers. When learning from repeated draws of signaling schemes, source-blind decision makers learn more slowly. However, as long as the average signaling scheme is $\varepsilon$ away from being completely uninformative, source-blind learning can match source-aware learning by using at most $O(1/\varepsilon)$ times more data.
Read the full abstract

All authors Nicole Immorlica, Brendan Lucier, Yaroslav Mukhin, Clayton Thomas, Ruqing Xu

Numerical Analysis

Finance-Informed Operator Learning for Option Pricing with Quantum-Compatible Realizations

Jiarui Feng, Bingyang Hu, Yu Jiang et al. (+2 more)

Pricing European options under local volatility requires repeatedly solving a PDE whose coefficients change with recalibration, while practitioners need both prices and sensitivities across spot-time surfaces. Neural surrogates can amortize these solves, but near expiry the solution loses regularity, making curvature difficult to learn and allowing violations of no-arbitrage bounds. We propose a finance-informed Deep Operator Network (FI-DeepONet) that decomposes the pricing operator into an input-dependent Black-Scholes carrier and a learned correction. The carrier uses strike-line integrated variance, capturing the leading near-expiry curvature singularity, while a smooth monotone Financial Admissibility Layer enforces pointwise price bounds. We derive a short-maturity estimate and differentiated asymptotics for the exact correction, together with identities linking correction error to price, sensitivity, and PDE-residual errors. On in-distribution tests, FI-DeepONet reduces global relative price error by nearly an order of magnitude versus vanilla and physics-informed DeepONet baselines. On parameter-shift OOD tests within the same local-volatility family, generalization is moderate. On index-option data, the frozen model remains accurate without market-specific retraining, though it does not outperform the same-input analytic formula. We also give a quantum-compatible realization in which selected linear maps are exactly compiled or approximated by a restricted diagonal-orthogonal family before supervised adaptation.
Read the full abstract

All authors Jiarui Feng, Bingyang Hu, Yu Jiang, Changhong Mou, Yeyu Zhang

Computation and Language

CoLMbo-SV: A Grounded Language Model for Explainable Speaker Verification

Massa Baali, Sarthak Bisht, Ziyue Qiu et al. (+3 more)

Speaker verification systems achieve high accuracy but provide little account of the acoustic evidence behind their judgments. Making these systems inspectable requires exposing interpretable evidence while retaining the richer information on which their decisions depend. We present CoLMbo-SV, a speaker language model that combines strong speaker discrimination with structured, acoustically grounded comparison reports. By connecting a pretrained speaker encoder to a language model and supplying explicit acoustic measurements, CoLMbo-SV makes voice comparisons inspectable without restricting verification to the evidence verbalized in its reports. We additionally introduce VoxReason, paired recordings with measured acoustic properties and comparison reports filtered through numerical and qualitative checks, providing supervision for this combined capability. We also develop an evaluation framework that separates what acoustic information a speaker representation encodes, what influences the verification score, and what the generated report discusses. On VoxCeleb1-O, CoLMbo-SV achieves 0.99% EER, reducing verification error by approximately 80% relative to the strongest audio-language baseline fine-tuned on VoxReason, while attaining a numerical-grounding score of 0.82. Our analysis further demonstrates that acoustic correctness and decision relevance are distinct properties of an explanation, exposing a gap that numerical-grounding metrics miss. Together, these contributions substantially advance audio-language speaker verification, bring its accuracy toward that of dedicated speaker encoders while adding checkable acoustic reporting, and establish an empirical framework for connecting natural-language explanations to the decisions they explain.
Read the full abstract

All authors Massa Baali, Sarthak Bisht, Ziyue Qiu, Joseph Konan, Rita Singh, Bhiksha Raj

Astrophysics of Galaxies

CHANG-ES XLII: Cosmic-Ray Electron Transport and the Spatially Resolved Radio-SFR Relation in Edge-on Galaxies

Jianghui Xu, Volker Heesen, Jiang-Tao Li et al. (+6 more)

Radio continuum emission traces star formation in galaxies, but cosmic-ray electron (CRE) propagation between the disk and halo complicates this connection. We present a wide-band, spatially resolved analysis of 19 nearby edge-on galaxies, combining LOFAR observations at 144 MHz with CHANG-ES VLA imaging at 1.575, 3.0, and 6.0 GHz, and hybrid H$\alpha$+22 $\mu$m star formation rate (SFR) maps. All maps are convolved to a common 2.1 kpc resolution for pixel-by-pixel analysis of the nonthermal spectral index $\alpha_{\mathrm{nth}}$ and the SFR surface density $\Sigma_{\mathrm{SFR}}$. All 19 galaxies exhibit statistically significant nonthermal spectral flattening with increasing $\Sigma_{\mathrm{SFR}}$, with a sample median slope of $0.19_{-0.03}^{+0.09}$ that traces the contrast between CRE injection in the star-forming disk and cooling in the halo. The galaxy-to-galaxy scatter in this slope shows no dependence on global galaxy properties and instead appears to be modulated by the local dynamical environment, with the largest $\alpha_{\mathrm{nth}}$-$\Sigma_{\mathrm{SFR}}$ slopes appearing in tidally perturbed systems. The resolved radio-SFR slope steepens from 0.57 at 144 MHz to 0.95 at 6 GHz, approaching linearity at high frequencies. This trend is consistent with energy-dependent CRE cooling and transport, with the deviation from linearity most pronounced at low frequencies, where the edge-on line of sight blends midplane and halo CRE populations. These results are robust against thermal contamination, beam smearing, and spectral curvature. Together, our analysis establishes $\alpha_{\mathrm{nth}}$ as an observational probe of projected disk-to-halo CRE transport in edge-on galaxies and quantifies how this transport reshapes the frequency dependence of resolved radio-SFR calibrations.
Read the full abstract

All authors Jianghui Xu, Volker Heesen, Jiang-Tao Li, Guilin Liu, Rainer Beck, Li-Yuan Lu, Yang Yang, Jayanne English, Shengtao Wang

Computer Vision

Background Gradients Shape Memorization in Flow Matching

Xuanhua Yin, Boyu Wei, Shuyi Zhang et al. (+3 more)

Repetition is closely associated with memorization in generative models, but how other training images affect the retention and copying of targets remains unclear. We study this question in class-conditioned flow matching, where images outside the target set form the background. At fixed target repetition and same-class background row count, replacing repeated same-class images with distinct images reduces the target extraction rate from 80.7% to 18.0%. To explain this effect, we develop a paired-trajectory framework that isolates target-induced parameter displacement and the background gradient response to it. This response has an exact path-integrated curvature representation, connecting background loss geometry to target learning. Reciprocal response transfer between repeated and distinct backgrounds changes target retention and copying in both directions, establishing the response's causal role. After target removal, the response correction parallel to the target-induced displacement preserves approximately 90% of the copying effects of full response transfer. Directly scaling the displacement also changes copying without further training. The post-removal copying effects of reciprocal transfer are reproduced across datasets and architectures. Together, these results identify the background gradient response as a mechanism through which same-class training data shape the retention of target learning and the reproduction of target images.
Read the full abstract

All authors Xuanhua Yin, Boyu Wei, Shuyi Zhang, Shunqi Mao, Chuanzhi Xu, Weidong Cai

Computation and Language

Beyond Calibration: Do a Typed-Decision Model's Probabilities Obey the Probability Axioms?

Keyi Li, Yihao He and Quanyi Li

Typed-decision models such as TypeSafe's Jev answer a declared yes/no or multiple-choice question about a state with a probability instead of text, and their evaluations report accuracy and calibration. Neither requires that the probabilities a model gives to logically related questions fit together item by item, which is what a system that acts on those probabilities needs. We test this property, coherence, with a battery of logically linked questions that needs no labels. For 160 items from ChaosNLI and PubMedQA, each with three mutually exclusive labels, we ask whether the label is X, whether it is not X, whether it is one of the other two labels, and which label applies. On 480 negation pairs, Jev's probabilities for "the label is X" and "the label is not X" miss summing to one by 0.064 on average (95% CI 0.055 to 0.072). Qwen3.8-27B, run from its official BF16 weights, misses by 0.293 with first-token probabilities and by 0.122 with verbalized probabilities. The gap to the first-token readout persists on pairs where both systems give similar probabilities, without double-negation labels, and after averaging Jev's repeated calls. Jev is not coherent either: its violations are about five times its repeat noise, and it over-endorses statements about single labels, so that its three single-label probabilities sum to 1.14 on average. The two systems also fail differently. Qwen3.8-27B's first-token readout under-endorses the complement of a label whether or not the question contains "not", rejecting both a statement and its negation in 196 of 480 pairs, and it does not become more coherent where it is more confident, whereas Jev's violations concentrate where its answer is uncertain. Because the checks need no labels, they expose biases that appear only when question forms are compared, and inconsistencies within items.
Read the full abstract

Artificial Intelligence

WorldAgent: Verification-Guided Agentic Physical World Construction

Caoliwen Wang, Mengdi Wang, Yige Chen et al. (+13 more)

Constructing complex physical worlds from language requires coordinating extensive 3D environments, detailed structures and objects at different spatial scales, and interacting physical processes under both stated goals and implicit physical constraints. We present WorldAgent, an agentic framework for verification-guided physical world construction from a single natural-language prompt, without iterative user debugging. A world construction layer expands the prompt into a structured world specification and uses physical knowledge to build scenes and run numerical simulations. After every step, a verification layer inspects scene geometry and simulation states alongside rendered views. Failed checks guide automatic revisions to the specification and re-execution of the affected steps. Accepted worlds pass the required checks and remain editable for further inspection and resimulation. We introduce AgenticSimBench, on which WorldAgent achieves the best scores among the evaluated agent-based methods on five of seven metrics. In a 26-participant user study, it receives the highest mean ratings across all four criteria.
Read the full abstract

All authors Caoliwen Wang, Mengdi Wang, Yige Chen, Zejia Wu, Bowen Huang, Siyuan Chen, Guanxiong Chen, Lifu Wei, Heng Zhang, Qinghai Zhang, Yin Yang, Guandao Yang, Shiying Xiong, Peng Wang, Chenfanfu Jiang, Peter Yichen Chen

Hardware Architecture

MorphAtt: A Neuromorphic Accelerator for Efficient Multi-Head Attention Processing in Spiking Vision Transformers

Rachmad Vidya Wicaksana Putra, Amirhesam Jafari Rad and Muhammad Shafique

Spiking Vision Transformers (SViTs) are developed as an energy-efficient alternative to conventional ViTs for computer vision tasks at the edge. However, huge parameter counts and complex multi-head self-attention (MHSA) operations make it challenging to achieve high energy efficiency in SViT inference, especially in tightly constrained applications. To maximize efficiency gains of SViT processing, we propose MorphAtt, a novel digital accelerator that expedites SViT inference through streamlined processing. Specifically, it processes MHSA operations using cascaded hardware modules: a Spiking Query-Key-Value generator (SpikeQKV), a low-complexity Spiking Multi-Head Self-Attention engine (SpikeAtten), and Reparameterization Convolution (RepConv) modules. To mitigate traffic congestion in on-chip memory accesses and data reuse, specialized inter-module buffers are integrated within the dataflow. Under synthesis using 32nm CMOS technology, MorphAtt achieves 792-1605 GOPS of throughput, while incurring ~39-55 mW of power consumption and 1.5 mm^2 of area, which lead to 20.3-29.1 TOPS/W of energy efficiency. These results also demonstrate that our MorphAtt offers better performance and efficiency trade-offs than state-of-the-art, thereby enabling highly energy-efficient vision-based AI systems at the edge.
Read the full abstract