Science & Mathematics

We present WorldStream, an open-ended 3D world model with three simultaneous capabilities: it extends playable terrain beyond any pre-specified map, acts on open-ended language in the running world, and restores visited places with their edits. A hierarchical planner over a typed decision model emits structured decision state: Macro fixes settlement, ecology, transport, and hydrological structure across a group of tiles; Micro specializes that contract into tile semantics under measured terrain and accepted neighbor commitments. Deterministic local code compiles accepted plans into terrain, roads, settlements, vegetation, and entities. Streaming separates their reach: plans run ahead along the predicted corridor, compilation only where movement will soon need it, geometry behind the user released. One world record retains plans, seeds, object identities and mutable state, boundary commitments, and edits by attribute, revised as the user advances and edits; returning tiles recompile from it. The same record grounds interaction: the model acts on existing entities, including stateful NPCs that update as they act, edits the present region, or attaches intent to unplanned regions. Together they yield a world expanding under bounded budgets, connected across independently planned tiles, and interactive.

Instruction-guided video generative models offer a scalable solution for stress-testing autonomous driving systems by simulating diverse environmental conditions. However, driving scenes operate within a dense Operational Design Domain (ODD) governed by strict traffic semantics. When applied to global driving-condition edits, general-purpose instruction-guided video editors frequently corrupt these semantics, leading to structural hallucinations and ``temporal shock.'' This work distills and protects traffic semantics during video editing at three levels. First, we construct a hybrid dataset by combining temporally aligned real-world videos with synthetic edits, balancing foundational scene structure with unambiguous appearance signals. Second, during training, we adapt advantage conditioning from offline reinforcement learning to align the model using dense rewards. We extract dense, pixel-level rewards from the hybrid data to serve as a localized signal and actively steer generation toward structurally reliable pixels. Finally, our model mitigates temporal shock by introducing thinking frames, a transition mechanism that seamlessly reconciles static image conditioning with dynamic video context. Extensive experiments demonstrate that our method prevents the corruption of multi-actor dynamics, establishing state-of-the-art spatiotemporal consistency, yielding consistent user preference (39% relative improvement in content preservation and 12% in temporal consistency), and achieving a significant 5% gain in downstream road line segmentation and 7% gain in downstream multi-object detection.

This study aims to deepen the understanding of artificial intelligence's (AI) impact on work by introducing and describing the AI-Job Dynamics Framework. Drawing from sociotechnical systems theory, complexity theory, and organizational learning theories, we conceptualize AI integration as a dynamic, non-linear process with three key phases. The framework, supported by theoretical propositions, identifies how this process affects organizational decision-making structures and key job characteristics over time, emphasizing the balance between centralization and decentralization of decision-making power. AI integration initially reduces job autonomy but may lead to greater autonomy in later phases as AI becomes more embedded. Throughout this process, AI reshapes skill requirements from routine to complex cognitive and adaptive skills. In the advanced phase of AI integration, we introduce the alternative paths and varied trajectories through which AI, organizational structures, and individual work experiences evolve. We conclude by discussing the implications of this framework for future research on AI and job design.

Cooke's Classical Model scores the statistical accuracy of experts by the $p$-value of a likelihood ratio test that compares the bin frequencies of realized seed variables with nominal probabilities, using a $χ^2$ reference. This calibration score drives performance-based weights and implicitly assumes independent seed variables, although seeds often share a common driver, such as the same indicator in several regions or periods. Under dependence the statistic converges to a weighted sum of $χ^2_1$ variables, and positive dependence makes the classical score anti-conservative. Under a Gaussian-copula working model, a Mehler-Hermite expansion gives the long-run covariance of the bin indicators in closed form, defines a design effect and an effective number of seed variables, and shows that the first-order correction coincides with the ad hoc calibration power of existing software. Rao-Scott-type and parametric-bootstrap corrected scores are proposed, together with a restricted maximum likelihood estimator of the nuisance dependence, computed from rank normal scores and invariant to location-scale miscalibration, and a prepivoted bootstrap that propagates its uncertainty. In simulations the classical score rejects up to 68\% of calibrated experts at a nominal 5\% level, whereas the corrected scores remove most of the excess rejection. In 45 published studies, design effects of 1.25-2 shift more than 5\% of the optimized weight in 29-44\% of studies. In three social-science studies with structured seeds, the baseline estimates imply that 18-21 seed variables carry the information of only 11-13 independent ones, although conclusions about individual experts depend on how the dependence is specified.

Inference-time scaling is often treated as a token-budget problem, but a failed coding agent can seek qualitatively different resources: computation, information, or feedback. We introduce Banzhaf Resource Attribution (BRA), a controlled factorial framework that measures each resource's marginal contribution to repair and its interactions while holding the failed state and revision protocol fixed. We evaluate six models on 194 function-level failures from MBPP+ and HumanEval+ and 126 repository-level failures from SWE-bench Verified. Task-relevant specifications consistently improve repair, whereas public-test feedback provides small gains and retrieved code little to none. The value of additional computation declines with model capability, while information remains useful; resource channels largely exhibit diminishing returns, although complementary information can reinforce itself. Yet three of four frontier models select feedback in 95-100% of cases and obtain lower gains than uniform random resource selection. BRA thus exposes a gap between the resources agents prefer and those that measurably improve repair.

Conventional visual sensing faces a fundamental dilemma: information that should remain private today may need to be recovered tomorrow. This challenge is becoming increasingly important as cameras become pervasive in our daily surroundings. Existing safeguards either protect visual content only after an image is formed or suppress information during capture in ways that prevent its later recovery. Here, we show through an information-theoretic analysis that privacy and recoverability cannot be achieved simultaneously in static optical systems. Information preserved in a measurement for later reconstruction also remains available to an observer of that measurement, limiting the privacy such systems can provide. To overcome this limitation, we introduce conditional imaging, an imaging approach that changes the optical state for every captured frame. The measurement alone conceals the scene, while knowledge of the corresponding physical system state enables its reconstruction. We analyze how the optical parameters determine the information that can be recovered and use this analysis to guide the design of a camera prototype. We evaluate the system on a captured dataset of human faces using increasingly powerful attempts to recover the scene, from direct visualization and computational inversion to learned reconstruction and identity recognition. Our experiments show that conditional imaging preserves identity in authorized reconstructions while withholding visual information from observers without access to the system state. By making the physical state of the camera part of the information needed for reconstruction, conditional imaging enables visual information to be protected at the point of capture while remaining recoverable for future authorized use.

This study proposes a sketch-to-simulation system to solve handwritten circuit diagrams. The system converts raw images into a Simulation Program with Integrated Circuit Emphasis (SPICE)-compatible netlist and automatically calculates the circuit's nodal voltages and currents. Another objective is to propose a holistic framework that benefits students and educators by enhancing the teaching and learning experience of electric circuits and expediting engineering design reviews. This article presents a multistage pipeline that converts raw photographs of handwritten electric circuits into executable SPICE-like netlists, providing a solution. A dataset of 1,000 hand-drawn Resistor-Inductor-Capacitor (RLC) circuits was curated and annotated for components, text regions, and character recognition. The proposed system decomposes a circuit diagram into three stages: (i) dual You Only Look Once (YOLO)-based detectors for components and text regions, (ii) a Convolutional Recurrent Neural Network (CRNN) with Connectionist Temporal Classification (CTC) and Optical Character Recognition (OCR)-based domain-aware symbol value extraction, and (iii) image processing with thresholding, skeletonization, and Density-Based Spatial Clustering of Applications with Noise (DBSCAN) to automatically identify nodes and map them with corresponding components. The proposed system produces the netlist and solves the circuit. The qualitative and quantitative ablation studies on modular stages, as well as end-to-end tests, demonstrate robust performance across various handwriting styles, lighting conditions, and clutter.This study serves as an interactive learning aid and practical tool for professionals, providing a comprehensive user interface (UI), enabling rapid digitization of circuit sketches for study, design, prototyping, or troubleshooting.

Digital twins of cerebral perfusion relate tracer dynamics to cerebral blood flow (CBF), blood volume (CBV) and mean transit time (MTT), but must be data-driven, physically consistent and uncertainty-aware. We introduce DuoPIC, a digital twin that needs no reference maps, in which a Windkessel flow model drives a tracer transport partial differential equation (PDE). A population-level neural closure divides each region into fast and slow pathways with tracer mass and MTT conserved by construction, and a tempered posterior quantifies the uncertainty of every estimate. Its voxel maps set the state of the art for lesion detection on ISLES'17 MR (AUC 0.745) and ISLES'24 CT (0.743). On held-out patients, the neural PDE reduced the compartment's reconstruction error by 19% and lifted 95% MTT interval coverage from 88% to 94%. Code: https://github.com/detinatran/duopic

High-quality human-translated parallel corpora can improve cross-lingual transfer, but are typically too small for large-scale pretraining. Machine-translated web corpora help overcome this limitation, making parallel data available at pretraining scale. Standard multilingual pretraining, however, samples source documents and their translations independently, without explicitly using their parallel structure. We investigate how to use the parallel structure of the data to improve multilingual pretraining by concatenating each document with its translation in the same training sequence. We compare this parallel pretraining (PPT) approach with standard pretraining (SPT), training models from scratch at matched token budgets. Surprisingly, despite the benefits of parallel data reported in prior cross-lingual work, PPT alone does not consistently improve language understanding over SPT. Our further analysis suggests that although PPT builds a strong cross-lingual foundation, as shown by its better use of shared information across languages, the mismatch between parallel training contexts and single-language task inputs limits its gains in general language understanding. This finding motivates a two-stage approach following the pretraining-fine-tuning paradigm: we first use PPT to build a cross-lingual foundation, then continue with SPT to adapt it to standard language modeling. In both high-resource and low-resource settings, the two-stage approach consistently improves general language understanding and cross-lingual task performance over standard pretraining alone at matched training budgets.

Early-grade reading screening is critical for identifying children at risk of reading difficulty. Existing AI-based screeners are often limited in the scope of their analyses and evaluation metrics, or are proprietary systems that lack transparent model specifications and evaluation. We present KIRA, a dual-pathway screening system that pairs word-level fluency scoring with detailed phoneme-level error detection. We adapt pretrained word and phoneme recognizers to children's speech, reducing word error rate from 13.88% to 8.84% and phoneme error rate from 41.66% to 23.00%. Powered by these child-adapted models, we implement Word Reading (WRE) and Oral Reading Fluency (ORF), two foundational tasks in K-2 screening. The system integrates these recognizers with a web-based teacher-facing interface presenting fluency scores alongside phoneme feedback for potential targeted intervention. KIRA brings modern child-speech recognition into a full-stack, deployable, classroom-ready reading screener.

This paper develops a spatial-frequency framework to quantify the impact of distributed energy resources with grid-forming inverters (GFMs) on voltage waveforms at different locations along a feeder. An analytical authority function is formulated to characterize how controllability varies with both electrical distance from the point of common coupling (PCC) and frequency. It is shown that the GFM control authority generally decays with both distance and frequency, reflecting the limited ability of inverter actions to influence voltage waveforms at distant points. To better capture controllability near resonance, a relative dominance metric is introduced to compare the influence of inverter-induced and grid-induced voltages at a given location and frequency. These results highlight that waveform controllability is fundamentally a spatial-frequency phenomenon along the feeder. Simulation studies validate the analytical findings by demonstrating both the spatial decay of control effectiveness along the feeder and the frequency-dependent trends in control authority in a realistic electromagnetic transient setting. The proposed framework provides insights into the placement and coordination of GFMs to enable effective waveform-level voltage control across electrical distance along the feeder.

Disturbance observers (DOBs) recover nominal behavior, but fast filters risk peaking. We propose a funnel-precompensated DOB for uncertain minimum-phase single-input single-output linear systems of known relative degree. A stable-factor residual yields a minimum-phase channel. At relative degree two, gain conditions require positive plant high-frequency gains and one positive common design gain; for higher degrees, the common-design result also uses a plant-gain interval and mismatch condition. Global boundedness, practical nominal-output recovery, and open-funnel initialization are established.

Visual question answering requires models to assign attributes and relations to the correct objects, a problem known as visual binding. Structured prompts make these assignments explicit through labeled fields. Yet an improvement from the full prompt does not reveal which labels help or how they interact. First, we isolate field-label effects across ten models and six tasks by holding the image and claim fixed. These comparisons reveal differences that overall accuracy can hide. With a relation label present, adding object-role labels helps Qwen3 but hurts Qwen2 on Spatial. Second, we compare auxiliary hints that preserve or swap the question's role assignments against a no-hint baseline. Across two image sets, Qwen3's advantage with preserving hints mainly reflects errors caused by swapped hints. Preserving hints remain near baseline. Field order further affects the harm. Third, we test clearer hint wording and direct selection of the relation shown in the image. Clearer hints improve text matching but do not consistently improve visual judgments. In contrast, relation selection partly reduces errors for some hints in Qwen3 and GPT-4.1-mini.

In this study we present a complete first-order Raman scattering spectrum of paramagnetic phase of MnTe$_2$- a material, where new type of spin texture induced by noncoplanar antiferromagnetic ordering below the Néel temperature ($T_N \approx 90$ K) gives rise to unconventional scenario of magnetism. Basic properties of all five phonon modes ($A_g$, $E_g$ and three $T_g$) predicted by the group theory have been successfully approached through polarization, excitation and temperature dependent measurements. In particular, the ever-missing phonon mode of $T_g$ symmetry has been unambiguously resolved in the room temperature Raman spectrum at approximately 107 cm$^{-1}$, just 1.5 cm$^{-1}$ higher in frequency compared to the $E_g$ mode. Density functional theory calculations reproduce the observed Raman-active phonon spectrum and support the assignment of all five first-order Raman modes, including the near-degeneracy of the $E_g$ and one of the $T_g$ vibrations. The calculations further reveal a dense manifold of low-energy antiferromagnetic spin configurations, separated from the ground state by only a few meV per formula unit. Furthermore, strong signals of inelastically scattered light from the paramagnetic (PM) phase of MnTe$_2$ were acquired with long wavelength excitations, i.e. 633 nm and 785 nm, revealing the $2A_g$ overtone at approx. 360 cm$^{-1}$. The enhanced absorption of the light in the near infrared range could be a putative reason behind the anomalous angular dependence of scattered intensity of the highest frequency $T_g$ phonon mode (183 cm$^{-1}$) measured in PM MnTe$_2$ with parallelly aligned polarizers. Finally, upon cooling just below the $T_N$, the measured properties of phonons remained basically unperturbed by the incipient long-range antiferromagnetic ordering.

Latent diffusion models almost universally place a Gaussian prior on the VAE latent space. For base generation this default is unproblematic; under reinforcement-learning fine-tuning, however, it carries a structural limitation. The standard VAE's KL term does not merely ignore mode structure - it actively taxes separation and tightness. When data are multimodal, the ELBO optimum compresses distinct modes into an overlapping latent. RL fine-tuning amplifies this overlap: the score network is under-resolved on the saddle region between modes, dynamic-to-variational amplification rises sharply, and the collapse boundary beta_c is pushed far above its variational scale. On the real alanine-dipeptide Boltzmann distribution, a Gaussian-prior VAE collapses at beta_c = 0.224 while a mode-tracked Gaussian-mixture prior collapses at 0.00316, a 71x gap from the same pipeline. We trace this gap to a single geometric mediator, d_min/(2 sigma_0), and verify the causal chain through controlled interventions: moving geometry at fixed prior, and switching prior at matched geometry (full mediation; prior-label coefficient 0.995x, 95% CI [0.79, 1.25]). A three-source decomposition localises the amplification to the learned policy (zero ODE-solver contribution, bounded anchor mismatch); the coverage plateau is an unconditional benefit of mode-separating priors that no amount of RL capacity recovers. The remedy is conditional: a non-Gaussian prior is necessary precisely when the latent geometry would fall into overlap and the task is coverage-sensitive. A machine-checked theory skeleton supports the KL pricing structure - a divergent compactness tax under a Gaussian prior versus a finite assignment fee under a mixture - and the position-versus-plateau distinction delimits what is finite-resource from what requires changing the prior.