Peer Revenue Publish or profit

Papers

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

All of arXiv

Saturday 10 October

Computer Vision

Prospective Forcing: Learning to Look Ahead for Autoregressive Long Video Diffusion

Yifan Shen, Yewei Xia, Gongxu Luo et al. (+6 more)

Autoregressive video generators synthesize long videos through chunk-wise rollouts, yet they are typically trained to predict only the immediate future. This limited prediction horizon makes it challenging to maintain video quality over extended rollouts. To address this problem, we introduce Prospective Forcing, a multi-head prediction framework that injects long-range predictive signals into autoregressive video representations. Built on a rolling-window generator, Prospective Forcing attaches lightweight, offset-specific heads that share the same conditioning information and independently predict chunks at different future temporal offsets. By extending supervision beyond the next chunk, these heads encourage the shared history representation to retain information useful for longer-horizon prediction, supporting temporally consistent generation. At inference time, we accelerate sampling by using accepted future-chunk predictions to skip full-window backbone executions. We equip the head with learnable embeddings to generate multiple candidates for the next chunk, and a learned verifier either selects a candidate or rejects all and falls back to standard autoregressive generation.
Read the full abstract

All authors Yifan Shen, Yewei Xia, Gongxu Luo, Peiyuan Zhu, Boyang Sun, Zhenhao Chen, Guangyi Liu, Guangyi Chen, Kun Zhang

Astrophysics of Galaxies

Multiphase Analysis of (U)LIRG Nuclear Activity (MAUNA). I. Spatially Resolved Optical Emission-Line Diagnostics

Justin A. Kader, Vivian U, Marina Bianchin et al. (+33 more)

We present resolved optical integral-field spectroscopy of the 32 luminous and ultraluminous infrared galaxy systems of the Multiphase Analysis of (U)LIRG Nuclear Activity (MAUNA) sample, drawn from the GOALS survey and selected to have JWST/MIRI medium-resolution spectrometer (MRS) observations. The data set combines new Keck/KCWI observations of 10 systems with archival VLT/MUSE cubes for 22 systems, covering H$\beta$ to [S II] at ~1 arcsecond spatial resolution (~0.9 kpc at the sample median distance). Using the optical [N II]/H$\alpha$ BPT diagram, we find that composite and LINER-like excitation together account for most of the classified area (median fractions 0.44 and 0.26), while Seyfert-like gas covers a median fraction of 0.02 but carries a far larger proportion of the line luminosity where it is present. Both the projected size of the Seyfert-classified region, $R_{\rm Sy}=0.08$-$7.5$ kpc, and its [O III] luminosity increase with the bolometric AGN luminosity, with $R_{\rm Sy}\propto L_{\rm AGN,bol}^{0.54\pm0.18}$, as expected for gas photoionized at a constant ionization parameter. The scatter about the luminosity relation depends on dust opacity: the [O III] luminosity per unit AGN luminosity falls by $1.09\pm0.25$ dex per unit depth of the 9.7 $\mu$m silicate absorption feature (Spearman $\rho=-0.60$, $p=0.001$). Optical excitation diagnostics therefore locate AGN-excited gas reliably but systematically underestimate the luminosity the AGN supplies to the host in the most obscured (U)LIRGs.
Read the full abstract

All authors Justin A. Kader, Vivian U, Marina Bianchin, Yiqing Song, George C. Privon, Sabrina Stierwalt, Tanio Diaz-Santos, Lee Armus, Vassilis Charmandaris, Ezequiel Treister, Rosalie McGurk, Hanae Inami, Jeffrey A. Rich, Sean T. Linden, Anne M. Medling, Loreto Barcos-Muñoz, Gabriela Canalizo, Aaron S. Evans, Kirsten L. Larson, Tianmu Gao, Thomas Lai, Laura Lenkic, Jaya Nagarajan-Swenson, Justin Howell, Cristina M. Lofaro, Matthew A. Malkan, María Sánchez-García, Christopher D. Martin, Mateusz Matuszewski, Claire E. Max, Joseph M. Mazzarella, James D. Neill, Nikolaus Z. Prusinski, Raymond Remigio, David Sanders, Jason Surace

Analysis of PDEs

Liouville theorems for stable solutions to the mean curvature equation

Fanheng Xu

We study stable solutions of the mean curvature equation $$ -\operatorname{div}\left(\frac{\nabla u}{\sqrt{1+|\nabla u|^2}}\right) =|u|^{p-1}u $$ in domains of $\mathbb R^n$. We establish Liouville-type results in three settings. First, for globally stable solutions in $\mathbb R^n$, we prove that $u\equiv0$ if $1<p<p_{JL}$, where $p_{JL}$ is the Joseph–Lundgren exponent. Second, for solutions in $\mathbb R^n$ that are stable outside a compact set, we show that if $n>2$ and $p=p_S$, where $p_S$ is the Sobolev critical exponent, then $u\equiv0$. Third, for solutions of the Dirichlet problem on connected unbounded domains with smooth boundary satisfying an epigraph-type geometric condition, we prove that if $n>2$, $1<p\le p_S$, and the solution is stable outside a compact set, then $u\equiv0$. The proofs are based on a gradient estimate for stable solutions, an integrability result, conservation laws and Pohozaev-type identities, and a unique continuation argument. These results extend Liouville-type theorems for stable solutions from the semilinear Lane–Emden equation to the mean curvature equation.
Read the full abstract

Computer Vision

TAG: Calibrating SAM3 for Universal Image Segmentation without Dense Supervision

Hao Zhu, Heyu Sun, Shuo Jin et al. (+2 more)

The recent Segment Anything Model 3 (SAM3) extends powerful promptable segmentation from spatial cues to concepts, raising the prospect of a single foundation model for universal image segmentation. However, this promise often fails to translate into reliable performance on downstream segmentation tasks. In this work, we empirically identify concept recognition, rather than mask quality, as the primary bottleneck: SAM3 accurately delineates the objects it recognizes yet struggles to robustly determine which concepts are present. Driven by this, we present \textit{\underline{T}ext \underline{A}nchor re-\underline{G}rounding (\TAG{})}, a simple yet effective framework that rectifies the concept recognition of SAM3 on the target domain without dense annotations. Specifically, \TAG{} introduces an Anchor Residual Pathway comprising image-conditioned, low-rank adapters to re-ground text anchors in the target distribution and modulate subsequent text-conditioned interactions. Guided solely by image-level category labels, \TAG{} jointly optimizes binary cross-entropy with Soft Boundary Ranking objectives to calibrate presence confidence and distinguish present concepts from confusable negatives. Building upon this, we devise Confidence-Weighted Distillation to couple rich semantic context of a multimodal large language model with fine-grained spatial priors of SAM3, thereby achieving effective calibration in the absence of manual annotations. Extensive experiments demonstrate that \TAG{} unlocks SAM3's segmentation potential with only 3.1M trainable parameters, achieving impressive performance across diverse scenarios, and notably, even surpassing leading mask-supervised methods on several metrics. \href{https://github.com/MiSsU-HH/TAG_SAM3}{Our code will be publicly available at GitHub \faGithub}.
Read the full abstract

All authors Hao Zhu, Heyu Sun, Shuo Jin, Jiayu Xiao, Feng Dai

Probability

Occupation Times at a Moving Interface and Sharp Yosida Rates for MacKean-Vlasov BSDE with Jumps and Fuzzy Selections

Kayembe Tshiswaka Tcheick, Mabela Matendo Rostin, Kabengele Mpunga Yannick et al. (+1 more)

We study a forward-backward McKean-Vlasov system with jumps whose backward equation contains a source obtained by evaluating the minimal-norm selection of a maximal monotone operator along the forward state. This selection changes branch across an interface that moves with the state law, the control, and an imprecisely known parameter described by fuzzy cuts. Assuming normal non-degeneracy and finite-statistics dependence of the interface on the law, we prove that the first and second moments of the time spent within a given distance of the interface are bounded, respectively, by a constant times that distance and by a constant times its square. The second-moment estimate makes it possible to integrate the source error before squaring it, which yields a Yosida error of first order in the regularization parameter in the natural solution norm; a first-moment argument gives only order one half. An explicit example shows that the first-order rate is sharp. For compact classes of admissible selections, the corresponding solutions form the cuts of a fuzzy solution set. These cuts are compact, stable with respect to the fuzzy level and the data, and approximated at the same first-order rate in Hausdorff distance. All estimates are uniform over a non-dominated family of probability measures representing volatility uncertainty. The quantitative results are obtained for separable quadratic potentials with finitely many affine branches.
Read the full abstract

All authors Kayembe Tshiswaka Tcheick, Mabela Matendo Rostin, Kabengele Mpunga Yannick, Bosonga Bofeki Jean-Pierre

Signal Processing

UL-Assisted Parametric CSI Feedback and Prediction for FDD Systems

Hao Chang, Mingcheng Nie, Shuangyang Li et al. (+4 more)

In frequency-division duplex (FDD) systems, the uplink (UL) and downlink (DL) channels are not fully reciprocal. Therefore, DL channel state information (CSI) estimated by each user equipment (UE) must be fed back to the base station (BS). However, because the channel varies between CSI acquisition and utilization, the feedback CSI may become outdated before it is used at the BS. To address this challenge, this paper proposes an UL-assisted parametric CSI feedback and prediction scheme for multi-user FDD systems with time-varying channel parameters. By exploiting the partial reciprocity between the UL and DL channel parameters, the BS first estimates the UL channel parameters from the sounding reference signals (SRSs) transmitted by the UEs. The BS then broadcasts DL pilots, allowing each UE to obtain observations containing information about its current DL channel. Instead of explicitly estimating the DL channel from these observations, each UE directly encodes them into feedback symbols and transmits the symbols to the BS. The BS decoder combines the feedback symbols with the partially reciprocal UL parameters to predict the future DL channel parameters. Finally, the corresponding DL channel matrix is synthesized from the predicted parameters using an explicit physical channel model. Simulation results on Sionna ray tracing (RT) dataset show that the proposed scheme outperforms the considered baselines in terms of channel prediction performance over different signal-to-noise ratios and prediction lengths.
Read the full abstract

All authors Hao Chang, Mingcheng Nie, Shuangyang Li, Chang Liu, Branka Vucetic, Yonghui Li, Wibowo Hardjawana

Mesoscale and Nanoscale Physics

Magnon Klein paradox

D. Wang

By considering the energy transport of spin wave, we show that, like the electron Klein paradox, there is no magnon Klein paradox for reflectivity. Heuristically, this can be understood according to the fact that, even if the ground state is unstable, any energy-carrying wave cannot have greater-than-one reflectivity off a passive interface, which will violate the conservation of energy. The magnon Klein paradox discussed in the literature is derived from assigning a wrong sign to the transmitted spin wave.
Read the full abstract

Computer Vision

Can Synthetic UAV Benchmarks Be Trusted? Real-Synthetic Validation for UAV Model Selection

Yuyi Zhi, Min Du, Bill Shi et al. (+4 more)

Synthetic UAV imagery can expand pre-deployment test evidence when task-specific Real data are scarce, but visual plausibility or semantic label alignment alone does not establish validity for model selection. We study when Synthetic evaluation preserves the model-selection conclusions obtained from Real UAV data through two complementary cases under fixed evaluation protocols. In a requirement-specific road-debris benchmark, four language-conditioned perception systems obtain lower absolute F1 on Synthetic data, yet the point-estimate ordering and top-ranked model are preserved; bootstrap analysis yields no supported pairwise reversal, and size-matched resampling preserves the top-model decision. In contrast, a closed-set stress test on a separately materialized business corpus, using four VisDrone-compatible detectors over eight empirically supported classes, changes the point-estimate top-ranked detector and reverses three of six pairwise point-estimate preference directions. A complementary audit of five UAV business scenarios further reveals substantial Real-Synthetic mismatch in class composition, object density, and target scale, including systematic under-representation of small targets. These results show that Synthetic UAV benchmark trust is conditional: semantic coverage is insufficient, and validity must be established empirically for the task, operating regime, and model panel of interest.
Read the full abstract

All authors Yuyi Zhi, Min Du, Bill Shi, Guodong Yuan, Junheng Liang, Peifan Liao, Xin Guan

Game Theory

Beyond Independent Values: Optimal Auctions with Correlated Valuations for LLM-Native Advertising

Shiman Sheng and Mingyu Guo

Conversational assistants are becoming the place where people ask what to buy, and sponsored content is moving into their answers. Several advertisers then compete for one sponsored slot, and all of them read the same conversation, so their willingness to pay usually moves together. Existing auctions for LLM advertising treat these values as independent. We therefore present AffMyerson, an auction that captures this correlation and turns it into revenue. We model the advertisers' values given the conversation as a multivariate normal and turn the economic premise of positive dependence (affiliation) into a sign condition on the precision matrix. We design a neural parameterization whose every output satisfies it. We prove that every prior in this family yields a truthful, ex-post individually rational auction, so estimation error can cost revenue but never truthfulness. We derive the conditional virtual values in closed form. By the design of the prior, pricing an auction needs no matrix inversion and takes O(n^2) time, fast enough to run at query time. We also work out the log-normal family used in empirical ad-auction work. There we prove that the auction is truthful for every prior and revenue-optimal when the prior is correct, and that the virtual value is regular everywhere whenever the conditional spread is below a loose threshold, which our fitted priors stay well under. With LLM-agent bidders, AffMyerson earns 6.2% more than a second-price auction with a tuned reserve, 6.1% more than Ronen's lookahead auction with the same prior, and 9.2% more than Myerson's auction run as if values were independent. This gain is still 4.5% on real keyword search queries where willingness to pay is hard to read. The correlation comes from common demand. Agents that judge the conversation bid together and agents told to calculate do not. The gain from conditioning moves with the correlation.
Read the full abstract

Atmospheric and Oceanic Physics

Learning native eddy-resolving ocean dynamics outperforms coarse emulation and downscaling

Zilu Liu, Tao Zhang, Hailong Liu et al. (+8 more)

Eddy-resolving ocean simulations are essential for representing mesoscale dynamics that regulate ocean circulation, energy transport, and climate variability. However, they are costly to generate, store, and share. Artificial intelligence (AI) may help, but it remains unclear whether high-resolution ocean fields can be faithfully represented by evolving ocean states at coarse resolution followed by spatial downscaling, or whether dynamical evolution needs to be learned directly from native high-resolution simulations. We show that directly learning dynamical evolution on a 0.1$^\circ$ grid provides a more faithful representation of mesoscale variability than evolving ocean states on a 1$^\circ$ grid followed by downscaling reconstruction. We developed a full-depth 1$^\circ$ emulator and a native 0.1$^\circ$ sea surface temperature (SST) emulator, both of which remained stable over long-term free-running simulations without significant drift, providing reliable baselines for this comparison. Direct learning at 0.1$^\circ$ preserves mesoscale energy, spectral characteristics, and frontal gradients, whereas the coarse-resolution evolution followed by downscaling approach recovers large-scale patterns but fails to reproduce key mesoscale signatures. We show that the learned mesoscale statistics remain consistent across diverse hardware platforms, supporting the reproducibility and portability of AI-based ocean simulations. These results suggest that downscaling of low-resolution large-scale states alone is insufficient to represent ocean mesoscale variability, supporting the need to learn dynamical evolution directly in a high-resolution state space. These findings highlight the continued importance of high-resolution physics ocean models as a foundation for AI-based ocean simulation, providing dynamically resolved training data for physically faithful high-resolution emulators.
Read the full abstract

All authors Zilu Liu, Tao Zhang, Hailong Liu, Kai Xu, Qinghui Lin, Maoxue Yu, Jiangfeng Yu, Jingwei Xie, Zhao Jing, Minghua Zhang, Lixin Wu

Artificial Intelligence

Shaping Latent Mixture-of-Experts: From Expert Geometry to Layerwise Capacity Allocation

Yaoxiang Wang, Kai Zhang, Ruobing Xie et al. (+4 more)

Fine-grained mixture-of-experts (MoE) models commonly increase expert count by narrowing each expert, treating expert shape as a by-product of routing granularity. LatentMoE moves expert computation into a lower-dimensional space, exposing the input dimension and intermediate width of each expert as explicit design choices. We show that an expert's local Jacobian rank is jointly bounded by these two dimensions, and that balancing them maximizes the rank ceiling at fixed per-expert parameters. Controlled shape sweeps across model scales and routing granularities consistently favor near-square experts. We further use layerwise intrinsic dimension (ID) to guide capacity allocation across depth, co-varying the number and shape of experts while keeping per-layer parameter count, activated FLOPs, and communication volume constant. Compared with layer-homogeneous baselines, our ID-guided allocation achieves lower validation loss and stronger downstream performance. These results establish expert shape and layerwise allocation as practical design axes for MoEs.
Read the full abstract

All authors Yaoxiang Wang, Kai Zhang, Ruobing Xie, Lanrui Wang, Guoliang Zhao, Junhui Li, Jinsong Su

Artificial Intelligence

AegisVox: Factorized Privacy and Clinical Preservation for Long-Form Speech

Praphul Singh, Akshat Agarwal, Ratul Ghosh et al. (+5 more)

Clinical speech reveals identity through both spoken identifiers and voice, yet its pauses, delivery, and speaker roles can also carry useful clinical evidence. Protecting that speech therefore requires more than preserving its words. We present AegisVox, a factorized framework that assigns consistent surrogate voices and edits identifier-bearing turns using document-consistent replacements. Across 390 public simulated or non-patient conversations totaling 63.6 hours, a matched comparison of two speech renderers separates replacement realization from retained identifiers, speaker linkage, and clinical preservation. Text-to-speech rendering realizes 535/601 planned replacements versus 453/601 for the codec-based speech editor. However, similar aggregate clinical preservation conceals more model-labeled contradictions, and conversation-wide speaker scores conceal greater edited-speech linkage under a score-direction sensitivity analysis. Fresh transcription of original audio provides a reference for interpreting the clinical measurement chain; an independent judge tests sensitivity to model-based assessment. Better replacement realization can therefore coexist with worse clinical or speaker-linkage measurements. AegisVox supports use-specific evaluation of those tradeoffs, not a guarantee of identifier removal or clinical safety.
Read the full abstract

All authors Praphul Singh, Akshat Agarwal, Ratul Ghosh, Aashna Kanuga, Devashish Khatwani, Anurag Dwarakanath, Sri Gadde, Krishnaram Kenthapadi

Quantitative Methods

Bayesian Uncertainty Quantification of Breast-Cancer Pathology AI Agreement under Reader- and Annotation-Condition-Dependent Reference Labels

Hammed A. Olayinka

Human reference labels used to evaluate medical AI can vary across readers and annotation workflows. We quantified agreement of a fixed pathology AI system with a future qualified reader without treating inferred consensus as error-free biological truth. Linked Evaluation and Unbiased NuCLS multirater data were analyzed using a fixed external HoVer-Net decision vector. The selected frame contained 1,144 anchors, 52 fields of view, five slides/patients, and five primary readers. The target was fixed-AI agreement with a future reader, with descriptive contrast $\Delta=\theta_E-\theta_U$. Analyses included raw agreement and iMRMC benchmarks, reader-only and shared-detectability Bayesian models, structural sensitivities, adversarial simulations, and predictive assessment. Among 5,188 reader-anchor pairs observed in both conditions, 19.62% changed AI-agreement state, while raw agreement was 0.5466 in Unbiased and 0.5473 in Evaluation. The reader-only model gave $\Delta=-0.0008$ (95% CrI, $-0.0033$ to $0.0008$), and the shared-detectability model gave $\Delta\approx0$ ($-0.0140$ to $0.0127$). Across structural perturbations, absolute agreement ranged from 0.420 to 0.646 in Unbiased and 0.409 to 0.640 in Evaluation. Frozen adversarial simulations met the comparative-stability criterion (median $R=0.295$; 14/16 model-scenario cells had $R<1$). Held-out-AI prediction supported shared predictive information, but varied by slide composition. The Evaluation-minus-Unbiased contrast was comparatively stable, whereas absolute agreement was structurally sensitive. These findings do not establish equivalence, causal assistance effects, biological accuracy, or new-patient transportability. Evaluations using variable human references should report reader population, selected frame, observation mechanism, and structural sensitivity.
Read the full abstract

Information Theory

Optimal Receive Design on the Observation Axis for RSMA Satellite-UAV Downlinks

Xianwei Meng

A receiver can change its observation state while the physical signal remains the same. In rate-splitting multiple access (RSMA) downlinks, this distinction separates receive-beam design and phase-reference transport from the motion of the satellite–UAV line of sight. We formulate receive optimization within the physical–observation dual-axis (PODA) structure and obtain three complementary results. For a known line-of-sight response and positive-definite receiver-noise covariance, a noise-weighted matched combiner simultaneously maximizes every decoding SINR and the minimum user rate. Under directional uncertainty, the principal eigenvector of the conditional response covariance globally maximizes mean array gain; for a Gaussian directional model, an equal-amplitude receiver is optimized by steering towards the conditional mean. Mean rate requires a different objective. We prove that the RSMA common-rate water level is jointly concave in the users' receive gains and derive a semidefinite relaxation with a computable global upper bound. A feasible rank-one beam can therefore be compared directly with the optimum of the finite-scenario problem. We also derive a precision-weighted linear predictor that suppresses uncertain physical-velocity estimates after observation-frame transport. Paired dynamic simulations reduce direction-cosine RMSE by 51.29% relative to unrestricted nine-point extrapolation. In a separate uncertainty experiment, direct rate optimization increases mean minimum spectral efficiency from 0.32425 to 0.41712 bit/s/Hz, an improvement of 28.64%; the corresponding design objective is within $2.28\times10^{-4}$ bit/s/Hz of a global upper bound. The phase-transport equation fixes a coherent representative of each optimal beam without changing its gain.
Read the full abstract

Optics

Near-focus intensity evolution of spatiotemporal optical vortex pulses and consequences for their nonlinear upconversion

Titouan Gadeyne, Vartika Vishnoi and Thierry Ruchon

Spatiotemporal optical vortex (STOV) laser pulses are attracting attention for their puzzling transverse orbital angular momentum (OAM) and propagation properties, which could determine their future applications. While the characteristic reshaping of their space-time envelope upon diffraction from the focal plane to the far-field is now well-established, the evolution of their intensity as they propagate has not been studied quantitatively, despite growing interest for their use in nonlinear optics and strong-field physics. Here, we explore experimentally and analytically the focusing dynamics of STOVs produced by a pulse shaper, and evidence a substantial longitudinal shift between the coordinate at which a circularly-symmetric STOV forms and that at which the pulse reaches its peak intensity. To illustrate some practical consequences for applications, we experimentally demonstrate how the yield of second-harmonic generation can consequently be maximized several Rayleigh lengths beyond the plane where the circular STOV really occurs. These results signal careful characterization of spatiotemporal pulses as a requisite for nonlinear optical experiments, and pave the way for finer control of complex laser drivers for light-matter interactions.
Read the full abstract

Systems and Control

Defending Consensus-Based Cooperative Sensing Against Conformity-Driven Byzantine Attacks in 6G UAV Swarm Networks

Ruixing Ren, Junhui Zhao, He Fang et al. (+4 more)

In integrated sensing and communication (ISAC) enabled six-generation (6G) networks, unmanned aerial vehicle (UAV) swarms increasingly rely on consensus-based cooperative sensing: each node iteratively exchanges and fuses decisions with its neighbors to approach a global picture of the environment, supporting the high-reliability, low-latency inference demanded by mission-critical tasks. Yet this very mechanism of conformity becomes the lever of Byzantine attacks. By exploiting the update rule, malicious nodes manufacture a false consensus that lures honest nodes into drifting away from their true observations round after round, so that a normal node's actual error probability evolves dynamically and rises far above its intrinsic sensing error. Conventional fusion models, which assume node errors are fixed and independent, are therefore severely mismatched. This article argues for a different defensive perspective: instead of hunting for bad nodes, we seek to understand how good nodes turn bad, characterizing the inducement dynamics with evolutionary game theory and handing this law to the fusion center so that it can dynamically correct the true error level of normal nodes. Simulations show that the approach sustains near-perfect situational inference across a wide range of attack intensities and network sizes, while naturally supporting low latency decisions. We further discuss several open challenges on the path toward secure ISAC-6G swarm sensing.
Read the full abstract

All authors Ruixing Ren, Junhui Zhao, He Fang, Xiaoke Sun, Jiamin Li, Pengcheng Zhu, Dongming Wang

Combinatorics

Generalized reflection complexity

Adam Blažek, Lubomíra Dvořáková and Edita Pelantová

The reflection complexity of a sequence, introduced recently by Allouche, Campbell, Li, Shallit, and Stipulanti, counts the factors of a given length up to reversal. We generalize this notion considering a group $G$ generated by antimorphisms over a given alphabet $\mathcal A$. The $G$-complexity of a sequence over $\mathcal A$ counts the factors of a given length up to the action of $G$. For $G$ generated by the reflection alone, the $G$-complexity coincides with the reflection complexity. We characterize eventual periodicity in terms of the $G$-complexity. Then we focus on the binary alphabet and the group $H$ generated by reflection and the antimorphism exchanging letters. We provide a simple formula for the $H$-complexity in the case of $H$-rich sequences and determine the precise values for complementary symmetric Rote sequences and generalized binary Thue-Morse sequences. Finally, we describe the aperiodic recurrent sequences having the minimum $H$-complexity – they are exactly the complementary symmetric Rote sequences, while the binary sequences with the minimum reflection complexity correspond to the Sturmian sequences.
Read the full abstract

Artificial Intelligence

PEARL: Process-Experience Aware Reinforcement Learning for Multi-Turn Agents

Jie Yang, Mianqiu Huang, Yiwei Zhao et al. (+3 more)

Terminal rewards make multi-turn agents trainable without step-level annotations, but group-relative advantages provide little guidance about which intermediate decisions to adjust. Failed trajectories also contain reasonable actions, while unfavorable decision patterns can recur as training proceeds. We introduce PEARL (Process-Experience Aware Reinforcement Learning), which improves local credit assignment without an additional critic. Given two continuations from the same history with different outcomes, PEARL locates their first divergence and builds a feature representation of each branch's local decision. It clusters these decision features in recent failure and success memories to identify recurring local patterns. Using this pattern memory, PEARL selectively strengthens negative advantages at lower-reward decisions, while leaving positive advantages and the base policy optimizer unchanged. On AppWorld, PEARL improves over ARPO on Normal and Challenge tasks at both 4B and 8B scales. On ALFWorld, it achieves 97.9% mean success at 4B, 2.7 points above the strongest baseline, while remaining competitive at 8B. Ablations of shaping location, success-bank filtering, and memory window support the value of selective, recurring-pattern-based updates.
Read the full abstract

All authors Jie Yang, Mianqiu Huang, Yiwei Zhao, Bo Wang, Yang Liu, Yuxin Wang

Computation and Language

PVFH-Bench: Benchmarking Premise-Induced Visual Fixation Hallucinations in Vision-Language Models

Xinhao Wang, Junbo Cui, Zhenghao Liu et al. (+4 more)

Large Vision-Language Models (LVLMs) have achieved remarkable progress across a wide range of multimodal tasks. However, object hallucination remains a persistent challenge, undermining model reliability in applications such as visual question answering, embodied agents, and human-AI decision support systems. In this paper, inspired by cognitive science studies on perceptual bias, we investigate a new failure mode termed Premise-Induced Visual Fixation Hallucination (PVFH). Specifically, PVFH refers to a premise-induced reasoning error in which models become anchored to visually confusable evidence, accept a false textual premise, and subsequently generate responses about an absent yet visually plausible object. To systematically study this phenomenon, we introduce PVFH-Bench, a benchmark that automatically constructs visually confusable but absent targets and pairs the core hallucination setting with diagnostic subsets to trace how models move from object recognition to Premise-Induced Visual Fixation Hallucination. Experimental results on 10 state-of-the-art LVLMs demonstrate that all evaluated models are highly vulnerable to Premise-Induced Visual Fixation Hallucination. Moreover, we find that explicit chain-of-thought reasoning can further exacerbate this failure mode by encouraging models to overinterpret weak visual cues as supporting evidence for hallucinated objects.
Read the full abstract

All authors Xinhao Wang, Junbo Cui, Zhenghao Liu, Yuan Yao, Xiaohua Li, Yu Gu, Ge Yu

Robotics

VbotNav-0: Asynchronous Cloud-Edge Dual-Policy Navigation across Indoor and Outdoor Environments

Shiqi Tan, Yujun Zhang, Keyu Li et al. (+4 more)

Continuous indoor-outdoor navigation connects destination-level reasoning with motion decisions that must respond to current observations. We present VbotNav-0, a cloud-edge dual-policy framework that links these processes through asynchronous policy-to-policy conditioning. The cloud policy produces subgoal guidance from visual history, navigation instructions, and route tokens; the edge policy combines available guidance with current RGB and LiDAR observations to generate local trajectories. This formulation allows the local perception-action loop to continue between cloud updates. A common route representation connects indoor paths and outdoor map routes, and the training recipe combines navigation demonstrations, corrective data, and grounded visual question answering (VQA). In the reported closed-loop evaluation of 258 episodes across 145 scenes, VbotNav-0 achieves 55.04% success rate and 55.03% success weighted by path length.
Read the full abstract

All authors Shiqi Tan, Yujun Zhang, Keyu Li, Shuaixi Pan, Siyu Zhou, Hailong Qin, Yinan Yu

Computer Vision

Automatic assembly modeling through attention-based assembly feature mate prediction and assembly feature pattern recognition

Ziheng Dai, Wanbin Pan, Zhihong Tang et al. (+1 more)

Assembly modeling plays an important role in constructing product models, yet it is still mainly performed via interactive operations that are experience-dependent and labor-intensive. Data-driven strategies hold promise for automating this process; however, existing approaches either predict poses that cannot directly generate editable B-Rep assembly constraints or require prescribed part pairs or predefined assembly faces. To reduce the large candidate space of constraint prediction without predefined assembly faces, we propose a data-driven framework based on assembly feature mate prediction. The framework performs automatic assembly modeling from unassembled B-Rep parts, using assembly features to group the geometric entities of local mating interfaces and restrict candidate formation. Given the lack of suitable assembly feature annotations in existing datasets, we construct a large-scale dataset through an automatic annotation pipeline. Using the annotated dataset, we train an attention-based network to predict assembly feature mates, with Engineering Compatibility Fusion (ECF) incorporating explicit engineering compatibility into learned matching scores. To improve the accuracy of automatic assembly modeling, we use assembly feature patterns (AFPs) to guide the matching process. Finally, a post-processing search selects assembly feature mate and AFP mate candidates, whose associated constraints are generated and solved to complete assembly modeling. Experimental results show that the framework achieves assembly constraint accuracies of 74.79% on the six-assembly comparison set and 64.09% on the extended test set of 16 assemblies, which includes the comparison set. With ground-truth face labels supplied as additional input, the corresponding accuracies increase to 81.78% and 80.93%, respectively.
Read the full abstract

All authors Ziheng Dai, Wanbin Pan, Zhihong Tang, Shuming Gao

General Economics

From Observation to Intervention: Mastering Bus Holding Control from Historical Uncontrolled Trajectories with Offline Reinforcement Learning

Minyu Shen, Xiaowen Bi, Weihua Gu et al. (+4 more)

Bus operations are inherently unstable—buses initially dispatched at equal headways will inevitably end up bunching up. This issue exacerbates passenger waiting times, creates load imbalances between buses, and undermines service reliability. Reinforcement learning (RL) has proven highly effective for mitigating bus bunching and outperforms traditional model-based approaches. However, existing RL approaches require agents to continuously interact with the environment, which potentially disrupts the bus system before effective policies are learned. This paper proposes an Offline Graph-based Multi-agent Actor-Critic (Offline-GMAC) method that trains bus holding policies using only readily available logged GPS data collected during normal uncontrolled operations, thereby eliminating the need for disruptive online interaction or exploration. The method employs general graph neural networks for multi-agent credit assignment. A key innovation of our method is a data augmentation scheme that synthesizes unperformed holding actions from historical uncontrolled bus trajectories. The scheme exploits the equivalence between stochastic link travel time delays and deliberate holding interventions. Under mild assumptions, we prove that the 1-Wasserstein distance between synthesized and target state transition distributions is bounded above by a term that scales linearly with the standard deviation of link travel times plus a tunable synthesis-dependent term. Evaluations using real-world bus data demonstrate that the proposed offline method achieves performance comparable to, and in some cases, even better than its online counterpart.
Read the full abstract

All authors Minyu Shen, Xiaowen Bi, Weihua Gu, Zhandong Xu, Yuezhong Wu, Feng Xiao, Lu Li

Information Retrieval

The Company a Question Keeps: Auditing Open Jev-Style Models for Retrieval

Jianbo Chen

Jev's documented interface evaluates typed questions independently against shared evidence. We audit request independence in seven open Jev-style checkpoints on 276 constructed three-passage states from 92 TREC retrieval queries. Paired interventions separately change requested questions, displayed evidence, and candidate eligibility. Request expansion induces strict binary flips in 9.1%/8.0% of Nimble's judgments and changes 21.4%/18.9% of selection-or-abstention actions in TREC DL2019/2020 under a threshold of 0.5 fixed before inference. Correct-action rates change by 0.0/+8.2 percentage points, respectively, with 95% query-bootstrap intervals of [-6.0,6.0] and [3.8,12.6]. Field masking traces all 70 request-set flips to the schema fields serialized into each scoring prompt. In Von, fixing a shared separator's position eliminates two eligibility-induced preference flips in selected diagnostic states. Laya-typed's default CUDA path produces five passage swaps without strict binary flips, and both yearly action-change intervals include zero; in an FP32 rerun, its probability movement vanishes at the model's four-decimal reporting resolution.
Read the full abstract

Computer Vision

DuSRFlow: Dual-Lens Super-Resolution via Parallax-Aware Flow Estimation and Complementary Reference Warping

Fangpu Zhang, Huanjing Yue, Zifan Cui et al. (+3 more)

Dual-lens super-resolution (DuSR) aims to enhance a wide-angle low-resolution (LR) image by exploiting a telephoto reference (Ref) image. A central challenge in DuSR is reliable reference utilization under large parallax, occlusion, and cross-camera imaging discrepancies. Existing methods often employ off-the-shelf monocular optical flow models to provide geometric priors. However, their assumptions of photometric consistency and smooth motion are violated in dual-lens scenarios, leading to unreliable alignment and degraded reference-based reconstruction. To address these issues, we propose DuSRFlow, a dual-lens super-resolution framework that improves reference utilization from both correspondence learning and reference aggregation. First, we introduce DuFlowNet, a dedicated dual-lens flow estimator for robust cross-view alignment. To train DuFlowNet without ground-truth flow annotations, we develop a real–synthetic data fusion strategy that composites synthetic foreground objects onto real dual-lens image pairs, generating controllable large-parallax samples while preserving realistic cross-camera imaging characteristics. Second, we observe that reference reliability is highly spatially uneven in DuSR, where unreliable local reference responses may suppress the utilization of reliable cues. We therefore propose a Complementary Reference Warping strategy that coordinates patch-based warping and kernel-free-guided deformable warping according to local reference reliability, producing a more reliable reference representation before LR–Ref fusion. In addition, we construct DuSR-RealV2, a challenging real-world benchmark with complex depth variations, large parallax, and occlusion. Extensive experiments on multiple DuSR benchmarks demonstrate that DuSRFlow achieves state-of-the-art performance in reconstruction quality, perceptual quality, and cross-dataset generalization.
Read the full abstract

All authors Fangpu Zhang, Huanjing Yue, Zifan Cui, Shaoxiong Tu, Qirui Yang, Jingyu Yang

Software Engineering

Executable but Unfaithful: Mock Object, Stub, and Assertion Generation by LLMs and Coding Agents

Hao Ning, Jiahui Dong, Zhen Yang et al. (+7 more)

Mocking is a foundational practice in unit testing, enabling developers to isolate the code under test by replacing real dependencies with simulated alternatives. Automating the generation of mock objects, stubs, and mock assertions has been a long-standing research goal. However, capture-and-replay approaches rely on recording the runtime behavior of actual dependencies. Evolutionary approaches such as StubCoder overcome this requirement but constrain generation to predefined symbol pools. Large language models (LLMs) and coding agents offer an alternative by generating mocking artifacts from source code without requiring recorded dependency executions or fixed symbol pools. Yet, no systematic study has evaluated their capabilities across all three artifact types. To address this gap, we construct MockBench, a benchmark comprising 165 mock artifact generation tasks from real-world Java and Python projects, with validated reference mock objects, stubs, and mock assertions. We evaluate six LLMs and three coding agents in terms of executability, behavior fidelity, and interaction-fault detection capability, complemented by comparisons with existing tools and sensitivity analysis. Experimental results show that LLMs and coding agents indeed have been the state-of-the-art techniques in mock-based testing. More specifically, compared with LLMs, coding agents substantially improve executability, while these gains do not consistently translate into higher behavior fidelity or better interaction-fault detection. Furthermore, we analyze 660 failures from LLMs and coding agents, and identify ten error types across the three artifact categories, revealing behavioral deviations that passing tests can conceal. Our findings highlight the challenges of mock artifact generation and provide practical guidance for both developers and researchers.
Read the full abstract

All authors Hao Ning, Jiahui Dong, Zhen Yang, Guanda Yan, Yichen Sun, Xuanyu Zhu, Zheng Shen, Yifei Pei, Jacky Keung, Jie M. Zhang

Number Theory

Examples of Bessel Functions of Order Zero Related to the Functional Equation of the Riemann Zeta Function

Hideto Iwata

In 2010, M. A. Chaudhry et al. introduced a class of functions Sigma satisfying an L^1-condition. They proved that every function phi in Sigma gives rise to the functional equation for the Riemann zeta function. They also provided an example of a function outside Sigma for which the functional equation still holds. In this paper, we present several examples of Bessel functions of order zero that do not belong to Sigma but nevertheless give rise to the functional equation for the Riemann zeta function. In particular, we show that J_0(a sqrt(x)), J_0(ax), and Y_0(x) exhibit this phenomenon.
Read the full abstract

Superconductivity

Chemical-Potential-Dependent Superconducting Spectroscopy on Bi2Te3/NbSe2 Heterostructures

Pengyu Hu, Chun Yu Wan, Qiaoyan Yu et al. (+21 more)

High-resolution scanning tunneling spectroscopy performed at dilution refrigerator temperatures was employed to investigate the superconducting properties of topological insulator/superconductor (Bi2Te3/NbSe2) heterostructures. Coherence peaks with inner kinks or inner peaks, deviating from the conventional single s-wave spectral line shape, were observed in both 5 quintuple-layer (5QL) and 7QL Bi2Te3 thin films. Based on the combination of chemical-potential-dependent spectroscopic measurements and model simulations, we conclude that these spectral features arise from contributions of bulk and surface states in Bi2Te3 by selectively proximity-coupled to a single superconducting gap from NbSe2, rather than inheriting the complicated superconductor's order parameter of NbSe2.
Read the full abstract

All authors Pengyu Hu, Chun Yu Wan, Qiaoyan Yu, Chen Qian, Yisheng Gu, Xian Du, Xuhui Ning, Tingwen Miao, Shi Chen, Zihao Ma, Hao Yang, Hemian Yi, Weijiong Chen, Xiaoxue Liu, Liang Liu, Dandan Guan, Yaoyi Li, Shiyong Wang, Canhua Liu, Sangen Zhao, Fulin Wang, Junwei Liu, Hao Zheng, Jinfeng Jia

Machine Learning

TagBO: LLM-Driven Task-Aware Graph Bayesian Optimization for Scientific Discovery

Xinzhe Yuan, Zimu Zeng, Zhuo Chen et al. (+6 more)

Mixed-variable optimization in scientific discovery often requires exploring large combinatorial spaces under costly experimental evaluations, making sample?efficient Bayesian optimization (BO) essential. However, existing BO methods typ?ically rely on structural descriptors to define similarity among categorical variables, implicitly assuming that structural similarity reflects similarity in task performance. In scientific optimization, this assumption is often violated, leading to surrogate representations that are poorly aligned with the optimization objective. To address this limitation, we propose TagBO (Task-aware Graph Bayesian Optimization), a method that integrates LLM-derived relational priors with descriptor-based struc?tural information for mixed-variable Bayesian optimization. Rather than directly using LLM outputs as predictive scores, TagBO treats LLM outputs as a potentially noisy source of task-aware relational priors and extracts coarse ordinal relational signals from LLM reasoning to construct task-aware similarity structures for scien?tific optimization. TagBO then performs constrained Gaussian process Bayesian optimization over the resulting joint discrete-continuous representation space. Ex?periments on diverse chemistry and materials formulation tasks demonstrate that TagBO consistently improves sample efficiency under fixed evaluation budgets across LLMs of varying capabilities.
Read the full abstract

All authors Xinzhe Yuan, Zimu Zeng, Zhuo Chen, Jianshu Zhang, Jinzong Dong, Jingyi Chen, Nanyang Ye, Huan Xiong, Qinying Gu

Artificial Intelligence

From Existing Methods to New Research Ideas: Two Perspectives on Recursive Self-Improvement

Leijun Zhou, Yanke Yu, Zhihao Liu et al. (+10 more)

Recursive self-improvement (RSI) offers a route to AI systems that participate in their own continued development. Recent methods allow language agents to retain changes to their parameters, memory, skills, and code for subsequent execution and improvement. We examine how these methods can inform new research ideas through two complementary perspectives. First, we introduce a unified taxonomy with nine design dimensions organized into four groups: update targets, update organization, update schedule and scope, feedback and selection. This taxonomy places existing methods in a shared design space, where examining individual dimensions and their interactions can inspire new research ideas. Second, we examine how existing RSI methods draw on traditional deep learning, for example through optimization, regularization, data augmentation, continual learning, and adversarial learning. We develop further ideas by examining more specific learning mechanisms and drawing on evolutionary computation. We also review how existing benchmarks assess self-improvement under online, offline, and offline-to-online settings. Together, these perspectives help researchers develop new RSI research ideas from existing designs and learning mechanisms. RSI papers and resources are available at https://prism-shadow.github.io/awesome-rsi/.
Read the full abstract

All authors Leijun Zhou, Yanke Yu, Zhihao Liu, Qu Xiang, Tiancheng Hu, Yiru Dong, Yihan Wu, Yangxi Xiang, Xuejun Wu, Xi Chen, Buyue Qian, Yaowei Zheng, Junhao Hu