arXiv:2607.00239v1 Announce Type: new Abstract: In 3D visualizations of natural phenomena, improving aesthetics can provide measurable benefits, but often involves transformations that affect how the data is perceived. As a growing range of tools - including AI-based methods - make visual design and modification more accessible, it is increasingly important to understand trade offs and concerns when making these changes. We conducted an expert survey (N=95) with visualization researchers, practitioners, and domain scientists, investigating reactions to fifteen alterations spanning presentation-level adjustments (e.g., lighting, camera position) and data-level modifications (e.g., removing errors, filling gaps), applied by both humans and AI systems. Results show differences in perceived acceptability are driven by the transformation's meaning, regardless of whether it operates at the presentation or data level. Additionally, certain modifications were consistently judged as more permissible than others regardless of human or AI authorship. While this relative ordering remains largely stable, AI-generated transformations are consistently rated as less acceptable than identical human-produced changes. These results reveal a distinction between more permissible and more sensitive alterations, and suggest the need for both designers and AI-assisted visualization tools to incorporate constraints and guardrails that reflect these differences.
Science Journals
arXiv:2607.00242v1 Announce Type: new Abstract: Corruption is embedded in networks of access, coordination, and protection, yet little is known about how gender shapes actors' positions within them. This article examines whether corruption networks in Colombia's territorial press reproduce gendered patterns of exclusion. Drawing on an access-to-power perspective, we argue that women's lower presence may reflect unequal incorporation into the spaces where exchanges are organized. Empirically, we use Transparencia por Colombia's Radiograf\'ia de Hechos de Corrupci\'on, integrating case- and actor-level information to build co-participation networks. We analyze gender differences in composition, position, recurrence, and institutional access to resources. Results show that these networks are strongly masculinized: men dominate actors and ties, and women appear less often among recurrent actors. However, women are not absent from connected or dense areas, suggesting uneven rather than absolute exclusion. The findings shift attention from whether women are ''less corrupt'' to how unequal access to power structures participation in corruption networks.
arXiv:2607.00244v1 Announce Type: new Abstract: For every $\varepsilon > 0$, Max Independent Set admits a polynomial-time $n^\varepsilon$-approximation algorithm on $n$-vertex graphs of effectively bounded twin-width [Berg\'e et al., STACS '23]. The approximation factor actually obtained is more precisely $n^{O(1/ \log \log n)}$. Prior to the current paper, no approximation hardness was known for this problem, and the existence of a polynomial-time approximation scheme (PTAS) was repeatedly raised as an open question. We answer this question in a strong sense: We show that there is a constant $\gamma > 0$ such that a polynomial-time $n^{\gamma/ (\log \log n)^2}$-approximation algorithm for Max Independent Set on graphs of twin-width at most 4 would refute the Exponential-Time Hypothesis (ETH). This lower bound further holds if a 4-sequence is provided as part of the input. We show the same hardness of approximation for Min Coloring, which also has a nearly matching $n^{O(1/ \log \log n)}$-approximation algorithm on graphs of effectively bounded twin-width. We also clarify the parameterized complexity of $k$-Independent Set on graphs of bounded radius-$r$ merge-width when the range of $r$ is limited. There is a fixed-parameter tractable algorithm for $k$-Independent Set on graphs given with radius-$2^{O(k^2)}$ merge sequences of bounded width [Dreier and Toru\'nczyk, STOC '25]. We complement this result by showing that $k$-Independent Set is W[1]-hard on graphs given with radius-$o(k)$ merge sequences of bounded width. We further show that this result also holds for $k$-Dominating Set.
arXiv:2606.25169v2 Announce Type: replace-cross Abstract: Sampling from an unnormalized target by reversing an Ornstein-Uhlenbeck diffusion requires the score of each noise-perturbed marginal. Tweedie's identity and a target-score identity give unbiased finite-reference estimators for this score. Scalar blends can reduce variance, but are too rigid for singular or strongly anisotropic targets. We cast blended score estimation as conditional risk minimization over matrix-valued blending coefficients, or gates, and derive the variance-optimal gate G*(y,t) = alpha_t^2 (alpha_t^2 I_d + gamma_t E[H_0(X_0) | Y_t = y])^{-1}, H_0 = -nabla^2 log p_0. Here alpha_t = e^{-t} and gamma_t = 1 - e^{-2t}. We call this formula the Laplace-Fisher Gate Identity (LFGI). Since the Tweedie-TSI disagreement has conditional mean zero, the gate changes estimator variance without changing its expected value. We give the Gaussian special case and prove finite-reference consistency and stability bounds for estimating the gate from weighted reference samples. We apply the finite-reference LFGI estimator to normalized density evaluation for Bayesian inverse problems. When MCMC pilot samples and derivative information are available, LFGI uses these byproducts to construct a normalized posterior-density surrogate. The surrogate enables posterior-energy evaluation, model-evidence estimation, and downstream density-based diagnostics. On a PDE-constrained inverse-problem benchmark, the LFGI surrogate improves posterior-density calibration and sampling diagnostics relative to the other tested score-estimator classes. Experiments using LFGI with known model evidence check absolute evidence calibration in both Gaussian and non-Gaussian settings.
arXiv:2607.01140v1 Announce Type: new Abstract: Open-vocabulary 3D Gaussian segmentation is challenging because it requires language understanding for diverse queries and accurate separation of Gaussians along object boundaries. Prior approaches either embed language knowledge into individual Gaussians to improve query responsiveness or optimize per-Gaussian instance features to encode object identity. However, these strategies may produce noisy Gaussian segmentations or rely on cost-inefficient per-scene optimization. We propose PairGS, a framework that reframes Gaussian segmentation as modeling pairwise relations between Gaussians. 3D Gaussian representations provide rich signals for relation estimation, such as view contribution weights and multi-view mask evidence. By leveraging these cues, PairGS explicitly constructs a relation graph for segmentation without a heavy optimization process. PairGS first proposes sparse edge candidates using low-dimensional descriptors, computes precise pairwise affinities only on those candidates, and builds a hierarchical cluster tree for multi-granular querying. It achieves state-of-the-art results on open-vocabulary 3D Gaussian segmentation benchmarks, while the fast variant is 50x faster than optimization-based instance-feature approaches.
arXiv:2607.01230v1 Announce Type: new Abstract: UAV swarms and cyber-physical multi-agent systems are increasingly deployed in safety-critical missions that require coordinated motion, distributed decision making, and autonomy. A major security risk arises when a legitimate agent is hijacked and driven by adversarial high-level commands. Rather than focusing on detection and isolation of malicious agents, we exploit a structural property common in autonomous platforms: low-level collision-avoidance modules are typically implemented as independent safety layers and may remain active even under high-level compromise. Building on this property, we propose a distributed containment framework that uses the compromised agent's uncompromised avoidance response as an indirect actuation channel. Defender agents select their geometric configuration to shape the repulsive field experienced by the target, with the goal of keeping it inside a prescribed admissible region and, when required, steering it toward a desired destination. The interaction is modeled as an online Stackelberg game in which defenders act as leaders and the adversary reacts by choosing the target command. Using support-function and normal-cone arguments, we derive an exact geometric characterization of robust one-step containment and introduce the notion of a repulsive cage. These results define a centralized Stackelberg oracle and motivate a fully distributed online approximation based on local communication and dynamic field estimation. We prove sublinear dynamic-regret bounds with respect to the centralized benchmark, quantifying the effect of network-induced estimation errors and temporal variability of the stage-wise optimum. Simulations validate the approach and corroborate the theory.
arXiv:2607.01144v1 Announce Type: new Abstract: While generative models have enabled training-free reward alignment, current methods typically excel in local exploration within narrow regions of the underlying distribution. These approaches struggle when preferences are unknown a priori and only revealed through sequential feedback-a scenario demanding broad exploration to uncover high-utility regions. To address this, we propose Sequentially-Controlled Interactive Multi-Particle Flow-Maps (IMPFM), a framework for sample-efficient online feedback-driven search. IMPFM progressively transports a group of interactive particles toward the target distribution, maintaining the broad coverage essential for heterogeneous preference alignment. IMPFM introduces a principled and efficient posterior sample sharing mechanism across particles powered by flow maps. By correcting individual particle drift with the collective posterior samples of the entire ensemble at each resampling step, the framework maximizes sample utility to enable global exploration while actively mitigating reward over-optimization, typical of standard control frameworks. Paired with a principled exploration-exploitation reweighting mechanism involving multi-particle interaction, this sequentially corrected multi-particle dynamics explicitly preserves structural diversity and overcomes the weight degeneracy inherent to standard SMC samplers. Crucially, we prove that the resulting sampling framework yields a multi-particle interaction-aware Feynman-Kac corrector that progressively steers the multi-particle system toward a KL-tilted target distribution, facilitating global exploration and preventing mode collapse. Extensive empirical evaluations and rigorous ablations across diverse search and alignment tasks confirm the efficacy of IMPFM over existing baselines.
arXiv:2607.00285v1 Announce Type: new Abstract: X-ray free electron laser (XFEL) experiments using self-amplified spontaneous emission (SASE) pulses typically achieve temporal resolutions of order several femtoseconds, as the pulse duration puts a practical limit to pump-probe or probe-probe schemes. Even with the emerging capabilities to generate pulses with attosecond durations with new single-spike SASE schemes, direct access to attosecond electron dynamics remains an experimental challenge. Here we show how X-ray coherent attosecond pulse-pair spectroscopy (X-CAPPS) provides a powerful new approach to access the ultrashort time-delay window. Coherent attosecond pulse pairs with time delays varying from ~500 as to ~5 fs are generated with Cu K$\alpha_1$ stimulated X-ray emission from a gain medium pumped by intense SASE XFEL pulses. These pulse pairs are analyzed with two subsequent Bragg crystal spectrometers, and the resulting interference spectrum is captured on two sequential 2D image detectors encoding their time separation, relative amplitudes, and phases with high precision. X-CAPPS requires no split-and-delay X-ray optics, nor XFEL pulse modifications, making it broadly implementable across existing facilities. This technique enables the investigation of attosecond processes with $\mathring{A}$ngstr\"{o}m resolution, providing a new tool for probing ultrafast dynamics across a wide range of atomic, molecular, and solid systems.
arXiv:2607.00286v1 Announce Type: new Abstract: Oscillatory neural networks (ONNs) have emerged as a promising neuromorphic architecture, leveraging coupled dynamical systems to perform computation and represent information through phase relationships. Their interactions can be designed to support intrinsic energy-minimizing dynamics, enabling tasks such as associative memory and optimization, and positioning them as a candidate architecture for continuous learning and inference. We present a neuromorphic primitive implemented using memristive edges with inhibitory couplings as a potential design for autonomous learning, and provide circuit simulation validation that the system is capable of denoising noisy inputs on an auto-associative task. While numerical Hopfield/Ising models routinely assume signed weights, neuromorphic implementations of ONNs often fail to realize negative weights due to device and circuit constraints. A practically implementable route to inhibitory (negative) weights is particularly valuable: it expands the class of attractor structures accessible to oscillator networks beyond purely synchronous couplings, and supports phase-coded memories where anti-phase constraints are not merely transiently enforced during training but can persist autonomously after release. We provide circuit simulations and theoretical analyses demonstrating that signed effective weights are necessary for anti-phase attractors to persist autonomously.
ADM-Fusion: Adaptive Deep Multi-Sensor Fusion for Robust Ego-Motion Estimation in Diverse Conditions
arXiv:2606.25111v2 Announce Type: replace Abstract: Robust multi-sensor fusion is essential for reliable autonomy in diverse and degraded environments, where sensor reliability can fluctuate rapidly. Because different modalities fail in distinct ways, effective fusion should adaptively balance complementary cues rather than rely on fixed weighting. This adaptability is particularly important for ego-motion estimation, since accurate updates depend on the consistent integration of complementary sensor information. We propose ADM-Fusion, an end-to-end deep learning based multi-sensor fusion method designed to adapt to environmental changes and sensor degradation. ADM-Fusion employs an adaptive sensor mixture-of-experts framework with content-aware routing to dynamically assign weights to sensor inputs in real time. The system further incorporates separate translation and rotation branches, coupled through a cross-task attention mechanism to preserve task-specific specialization while enabling information sharing. ADM-Fusion is trained on the CARLA-LOC simulated dataset and subsequently fine-tuned on KITTI real-world data, demonstrating effective simulation-to-real transfer. Experiments show that ADM-Fusion remains robust under degraded conditions while maintaining competitive performance against existing methods.
arXiv:2607.00292v1 Announce Type: new Abstract: Designing deployable and resilient network topologies from natural language requirements remains a challenging problem in network automation. This work investigates the ability of Large Language Models (LLMs) to generate structurally valid and constraint-compliant network topologies through a constraint-driven pipeline combining hierarchical modeling and systematic validation. The framework is evaluated via a multimodel comparison of proprietary and open-weight LLMs across four realistic network scenarios released as a public dataset. We assess structural correctness using node and edge F1-scores against reference topologies, and evaluate resilience through server and content connectivity metrics. In addition, we analyze common failure modes, including interface mismatches and directional inconsistencies in generated topologies. Overall, this work provides a systematic benchmark for understanding how LLMs handle structural and resilience constraints in topology synthesis, and supports informed model selection for AI-driven network design.
arXiv:2607.00296v1 Announce Type: new Abstract: Human motion forecasting in unconstrained real-world videos remains challenging due to the ambiguity of future behaviors and the presence of noisy multimodal observations. While facial affect potentially provides complementary behavioral cues, its practical utility and mechanistic boundaries within motion forecasting frameworks remain poorly understood. In this work, we present a systematic study investigating the utility and temporal limitations of affect-conditioned forecasting in-the-wild. We establish a rigorous multimodal pipeline combining MediaPipe body pose trajectories with HSEmotion facial affect representations, and introduce the Gated Affect Transformer (GAT) to dynamically regulate cross-modal information flow. Through extensive multi-horizon evaluations under a strict subject-wise protocol, we demonstrate that naive early cross-modal concatenation consistently degrades forecasting accuracy relative to pose-only baselines. Conversely, our proposed gating mechanism stabilizes cross-modal integration by adaptively controlling the affective stream. Crucially, controlled counterfactual experiments using shuffled and randomized affect inputs reveal that the learned gate successfully suppresses unstructured cross-modal noise while remaining responsive to plausible affective signals. Furthermore, our empirical results indicate that facial affect features provide bounded, horizon-dependent predictive cues strictly within short-to-medium windows (e.g., 30 frames), whereas long-term trajectories remain predominantly governed by intrinsic kinematic continuity. Our findings provide empirical evidence that facial affect should be regarded as a complementary behavioral cue rather than a dominant driver of future motion, offering practical guidance for selective multimodal fusion in unconstrained human motion forecasting.
arXiv:2606.25997v2 Announce Type: replace Abstract: There has been significant recent progress in algorithms for approximation of Nash equilibrium in large two-player zero-sum imperfect-information games and exact computation of Nash equilibrium in multiplayer strategic-form games. While counterfactual regret minimization and fictitious play are scalable to large games and have convergence guarantees in two-player zero-sum games, they do not guarantee convergence to Nash equilibrium in multiplayer games. Recently, an approach has been presented for exact computation of Nash equilibrium in multiplayer imperfect-information games that solves a quadratically constrained program based on a nonlinear complementarity problem formulation derived from the sequence-form game representation. This formulation was solved using Gurobi's nonconvex quadratic solver, which employs spatial branch-and-bound to iteratively refine variable bounds by solving convex relaxations of bilinear terms via McCormick envelopes. During presolve, Gurobi introduces auxiliary variables and, in some cases, binary variables, leading to an internal MIQCP reformulation. This approach was demonstrated to outperform prior algorithms from the Gambit software suite and quickly solve three-player Kuhn poker after removal of dominated actions; however, the algorithm was not able to solve the full version of the game within 24 hours. In this paper, we derive finite bounds on slack and multiplier variables in the nonlinear complementarity formulation. These bounds strengthen the convex relaxations used within spatial branch-and-bound and lead to substantial computational improvements. We demonstrate the impact of the proposed bounds on exact Nash equilibrium computation in three-player Kuhn poker.
arXiv:2607.00669v1 Announce Type: new Abstract: Autoencoders (AEs) have emerged as powerful tools for non-linear dimensionality reduction, often surpassing traditional linear methods such as Proper Orthogonal Decomposition (POD) in scenarios characterized by slowly decaying Kolmogorov $n$-widths. In the realm of Reduced-Order Modelling (ROM), these models are increasingly utilized to learn low-dimensional representations of solution manifolds associated with parametric Partial Differential Equations (PDEs). However, the high expressivity of AEs presents a challenge: although trained networks typically minimize reconstruction error, they often struggle to capture the essential properties necessary for building accurate and robust ROMs. Recent works by arXiv:2307.15288v2 and arXiv:2506.11641v1 have tackled this challenge in fully connected AEs by proposing representation-consistent architectures, which preserve some of the properties belonging to POD. This study builds upon that concept by extending representation consistency for convolutional layers. We introduce a novel class of symmetric Convolutional AutoEncoders (CAEs) designed to embody the primary properties of manifold parametrization mappings. When integrated into a ROM framework, this architecture demonstrates significantly improved predictive capabilities. Specifically, we compared the performance of the ROMs based on classical and symmetric CAEs on three one dimensional academic test cases, namely the Linear Advection, the Viscous Burger and the Kuramoto Sivashinsky equation. Numerical results demonstrate that our proposed symmetric approach consistently yields more accurate latent trajectories, lower reconstruction errors, and enhanced model robustness.
arXiv:2607.00300v1 Announce Type: new Abstract: We study price-adjustment dynamics for computing competitive equilibria (CE) in Fisher markets with chores. Unlike in classical goods markets, prices in chores markets are payments for taking on undesirable tasks, and natural excess-demand dynamics can fail; even the na\"ive analogue of Walrasian t\^atonnement may diverge. Recent work of Chaudhury et al. [2025] overcomes this obstacle via relative t\^atonnement, which subtracts the average excess-demand signal from the excess demand vector. This recovers convergence, but at the cost of coupling the price updates across all chores. This leaves open whether such global coupling is inherent, or whether convergent t\^atonnement can be recovered through a genuinely local update in which each chore reacts only to its own excess demand. We answer this question affirmatively through multiplicative t\^atonnement, a fully distributed dynamics in which each chore price is updated using only its current price and its own excess-demand signal. Although the update contains no explicit normalization term, Walras' law and the multiplicative form of the update implicitly preserve the relevant aggregate price geometry. We prove that multiplicative t\^atonnement converges to a CE in any chores Fisher market with continuous, convex, and $1$-homogeneous (CCH) disutilities. For convex CES disutilities, we further prove an approximate-CE convergence rate with the same $O(1/\varepsilon^2)$ dependence as relative t\^atonnement, but with improved dependence on problem constants. Experiments on real-world and simulated instances show that multiplicative t\^atonnement is substantially faster in practice, often by an order of magnitude.
arXiv:2607.00301v1 Announce Type: new Abstract: The emergence of powerful deep generative models based on diffusion and flow matching has enabled the learning and modeling of complex distributions. Learning quantum distributions, however, remains challenging due to the inherent difficulty of accurately modeling the meaningful physical properties of quantum states. We propose Quantum Flow Matching (QFM), a novel generative model designed to learn quantum distribution by utilizing spin Wigner function and flow matching. By converting density matrix into the spin Wigner function and leveraging functional flow matching to learn distributions in function space, QFM enables accurate and effective learning of multi-qubit quantum distributions. We demonstrate the effectiveness of our method by evaluating physical quantities such as trace, purity, and entanglement entropy of the generated quantum states, accurately capturing the underlying physics of the given quantum distributions.
arXiv:2607.00912v1 Announce Type: new Abstract: This paper studies the stochastic performance of a downlink multiple-input multiple-output integrated sensing and communication (ISAC) system over Rician fading channels. Rician fading is important in line-of-sight (LoS)-dominated deployments, where a deterministic propagation component can strongly affect sensing and communication reliability. The base station (BS) simultaneously serves a user and senses a target. The BS-user channel contains LoS and non-line-of-sight components. The user LoS angle may be fixed or random, and the target angle may follow an arbitrary distribution potentially correlated with the user angle. Compared with Rayleigh fading, the deterministic LoS component introduces angle-dependent terms and leads to generally independent but non-identically distributed random vectors, requiring new analysis. We analyze two beamforming strategies: subspace joint beamforming (SJB), optimal for the shared waveform structure, and linear beamforming (LB), a practical alternative using separate sensing and communication beamformers. For both schemes, we derive communication outage probability (OP) and sensing OP based on the Cramer--Rao bound (CRB). We also identify special cases with simpler expressions. For LB, we derive upper and lower bounds on sensing OP and a tractable approximation. We characterize large-system and high-power scaling laws. LB without dirty paper coding (DPC) is interference-limited at high power due to radar self-interference. Results show the Rician K-factor affects communication more strongly than sensing, with non-monotonic behavior across regimes. LB with DPC achieves the best overall performance in strong LoS environments and is the only scheme achieving ultra-high communication reliability in Rayleigh fading, while SJB provides a robust lower-complexity alternative across operating conditions.
arXiv:2607.01179v1 Announce Type: new Abstract: Scaling inference compute, by generating many parallel attempts per problem, is a costly but reliable lever for improving language model capabilities. By default these attempts are generated independently, wasting inference compute on redundant solutions. This waste seems unavoidable. After all, independence is what makes parallel sampling trivial to scale. However, this tradeoff is not fundamental: there is a rich design space of samplers that generate correlated but exact samples entirely in parallel. We explore this design space as an avenue for improving sample efficiency in scaling inference compute and reinforcement learning (RL). Concretely, we introduce QuasiMoTTo, which uses correlated samples as a drop-in replacement for i.i.d. samples. To generate these samples, QuasiMoTTo uses a reparameterization of autoregressive sampling as inverse-CDF sampling and draws the underlying uniforms with quasi-Monte Carlo (QMC); because QMC spreads the uniforms out more evenly than i.i.d., the resulting samples cover the output space with far less redundancy. Even though the batch is correlated, each sample is marginally distributed according to the language model, so we can use the batch for policy-gradient training. Our empirical analysis focuses on understanding how efficiently QuasiMoTTo can turn compute into performance. To evaluate correlated samplers, whose dependence breaks standard pass@k estimators, we first develop an unbiased bootstrap estimator. Across four reasoning benchmarks, QuasiMoTTo matches i.i.d. pass@k accuracy with 25-47% fewer samples. Strikingly, QuasiMoTTo often saturates an upper bound on pass@k that holds for any marginal-preserving sampler. We also apply QuasiMoTTo to policy-gradient RL (GRPO) where it matches i.i.d. performance with 50% fewer training steps. These gains come from higher coverage, which yields a stronger learning signal per batch.
arXiv:2607.01202v1 Announce Type: new Abstract: We present World from Motion, a method for generating freely renderable dynamic 3D Gaussian representations from monocular videos. Our approach conditions a video model on dense, pixel-aligned renderings that encode appearance, geometry, and 3D scene motion along both input and target camera trajectories to correct rendering artifacts and fill in missing regions from an initial reconstruction. To train this model, we construct a dataset of aligned multiview video pairs and dynamic 3DGS representations, with simulated artifacts characteristic of monocular reconstruction. At test time, we distill the model's generations, including newly observed regions and motions, back into a single consistent, high-quality dynamic 3DGS, improving both novel-view synthesis and the underlying 3D motion. Our method sets a new state of the art in 4D reconstruction and seamlessly generalizes to in-the-wild videos with large viewpoint changes and dynamic motions.
arXiv:2607.01225v1 Announce Type: new Abstract: Prior work on imitation learning from suboptimal demonstrations typically relies on compressed supervision signals such as confidence estimates, discriminator scores, or importance weights. These scalar signals are inherently limited, as they cannot explicitly express intermediate reasoning about task progress, failure modes, or corrective actions. We propose a language-critique framework for imitation learning from suboptimal demonstrations that instead leverages natural language as a structured supervision signal, avoiding the collapse of expressive feedback into scalars. Our method first constructs language labels from demonstrations that explicitly describe current progress, identify suboptimal behaviors, and provide fine-grained corrective guidance. We then introduce a language-critique loss that directly trains policies using these structured signals without reducing them to scalars, and instantiate it for both behavior cloning and diffusion policies, yielding LC-BC and LC-DP. We further provide a theoretical result showing that the proposed objective upper-bounds the expert performance gap under standard assumptions. Empirically, we evaluate on diverse continuous control tasks spanning navigation, manipulation, and gameplay, where our methods consistently outperform strong imitation learning and offline reinforcement learning baselines. These results demonstrate that language can serve as a powerful and structured form of supervision for learning robust policies from suboptimal data.
arXiv:2607.01233v1 Announce Type: new Abstract: LLMs are increasingly used to brainstorm research ideas, but existing evaluations mostly judge individual ideas by novelty, feasibility, or expert preference. We instead ask: how far are current LLM-generated ideas from human researchers? To characterize this gap, we build a large-scale evaluation framework for ideation from high-quality human research papers. For each paper, we reverse-engineer a small set of closely related prior works that likely inspired its core idea. LLMs are then prompted to generate a new idea from the set of paper titles and summaries. We introduce a two-axis research-taste taxonomy to profile each idea by its opportunity pattern and research paradigm, and use it to quantify the divergence between human and LLM ideas. Across idea sets generated by different LLMs, we observe a consistent distributional gap: LLM ideas are disproportionately concentrated around bridge-like opportunities and synthesis methods, whereas the human paper reference distribution spreads more broadly across ways of framing gaps and constructing contributions. This result suggests that strong LLMs can produce a range of reasonable ideas, but that range remains narrower than, and systematically shifted relative to, human research taste.
arXiv:2607.00335v1 Announce Type: new Abstract: We provide rigorous error analysis of the mass-preserving time-splitting methods for solving the semiclassical Dirac equation. The scaled Planck constant $\epsilon$ in the equation gives rise to rapid oscillations in both space and time when $0<\epsilon\ll 1$ with wavelengths of order $O(\epsilon)$. %We prove that the first-order splitting $S_1$ and the second-order splitting $S_2$ schemes preserve the total discretized mass. Rigorous error estimates reveal the precise dependence of the approximation errors on the time step $\tau$, the spatial mesh size $h$, and the parameter $\epsilon$. Specifically, the temporal error scales as $O\left(\tau/\epsilon^2\right)$ for the first-order splitting $S_1$ and as $O\left(\tau^2/\epsilon^3\right)$ for the second-order splitting $S_2$, while the spatial error scales as $O(h^m/\epsilon^m)$ for both methods, where $m$ is related to the regularity of the solution. In addition, we obtain error bounds for key physical observables, including the total probability density $\rho$ and the current density $\mathbf{J}$. Compared with finite difference time domain (FDTD) methods, time-splitting approaches exhibit spectral accuracy in space and retain a relatively low computational cost. Furthermore, we demonstrate that higher accuracy can be achieved by employing the fourth-order compact time-splitting ($S_\text{4c}$) method. Numerical experiments are conducted to verify the reliability of the error estimates.
arXiv:2607.00338v1 Announce Type: new Abstract: Object detection for Unmanned Aerial Vehicles (UAVs) working in open and dynamic environments is a highly challenging task. While Vision-Language Models (VLMs) have offered a powerful solution for universal object detection, adapting them to UAV scenarios remains non-trivial due to a substantial domain gap between VLM pre-training data and aerial imagery. The prevailing Parameter-Efficient Fine-Tuning (PEFT) methods prove ineffective in bridging this gap, as VLMs' "natural-scene, foreground-dominant" visual priors misalign with the "bird's-eye-view, background-dominant, small-object" characteristics of UAV data. To address this issue, we propose DroneFINE, a novel PEFT paradigm comprising two domain-aware complementary modules tailored for VLM-based drone image detectors. Specifically, a data-dependent, foreground-aware, and multi-path adaptation mechanism named HyperAdapter is designed, which overcomes the static structural constraints of PEFT. In addition, a background suppression algorithm named SemanticGate is developed. It is a text-conditioned guidance strategy that employs background vocabulary to actively guide the model in suppressing responses from irrelevant regions. Extensive experiments on VisDrone and UAVDT demonstrate that DroneFINE significantly outperforms existing PEFT methods and achieves performance comparable to full fine-tuning while substantially reducing the number of trainable parameters.
arXiv:2607.00253v1 Announce Type: cross Abstract: Dynamic metasurface antennas (DMAs) enable programmable wave-domain signal processing that can be jointly optimized with downstream digital processing in an end-to-end manner. Existing studies, however, typically assume ideal analog-to-digital conversion (ADC) and often rely on simplified electromagnetic models. Here, we study ADC-aware end-to-end optimization of a monostatic sensing pipeline based on a DMA with strong mutual coupling (MC). We model the wave domain using an MC-aware multiport-network model whose parameters were experimentally estimated for a fabricated chaotic-cavity-backed DMA with 96 one-bit-programmable meta-elements. We perform ADC-aware end-to-end optimization of the DMA configurations and digital classifier, either with awareness of a fixed uniform ADC or, optionally, with jointly learned ADC decision thresholds, and compare against baselines that assume an ideal ADC and/or ignore MC. Our results show that ADC awareness is essential in low-resolution ADC regimes: with one-bit ADCs and eight DMA configurations, deploying an ideal-ADC-trained system with a uniform one-bit ADC reduces the test accuracy from 95.5% to 56.0%, whereas ADC-aware training with the same fixed uniform one-bit ADC achieves 87.2%. We also show that without MC awareness the accuracy drops to the random-guess level. Learning non-uniform ADC thresholds provides at most modest additional gains over fixed uniform ADCs in the considered DMA-based sensing pipeline.
arXiv:2603.16859v2 Announce Type: replace Abstract: Omni-modal large language models (OLMs) redefine human-machine interaction by natively integrating audio, vision, and text. However, existing OLM benchmarks remain anchored to static, accuracy-centric tasks, leaving a critical gap in assessing social interactivity, the fundamental capacity to navigate dynamic cues in natural dialogues. To this end, we propose SocialOmni, a comprehensive benchmark that operationalizes the evaluation of this conversational interactivity across three core dimensions: (i) speaker separation and identification (who is speaking), (ii) interruption timing control (when to interject), and (iii) natural interruption generation (how to phrase the interruption). SocialOmni features 2,000 perception samples and a quality-controlled diagnostic set of 209 interaction-generation instances with strict temporal and contextual constraints, complemented by controlled audio-visual inconsistency scenarios to test model robustness. We benchmarked 12 leading OLMs, which uncovers significant variance in their social-interaction capabilities across models. Furthermore, our analysis reveals a pronounced decoupling between a model's perceptual accuracy and its ability to generate contextually appropriate interruptions, indicating that understanding-centric metrics alone are insufficient to characterize conversational social competence. More encouragingly, these diagnostics from SocialOmni yield actionable signals for bridging the perception-interaction divide in future OLMs.