arXiv:2607.05563v1 Announce Type: new Abstract: Interpretable explanation methods in Artificial Intelligence aim to uncover the underlying causes and their effects, enabling a deeper understanding of why a system behaves in a certain way under different inputs. Unlike traditional explainability methods, which mainly highlight correlations between input and output variables, causal explanation focuses on interventional questions. By doing so, it provides more robust insights, helping users understand automated decisions, especially in high-risk domains. Recovering an explicit directed causal structure, however, is often impractical in large-scale, hybrid cyber-physical systems with feedback loops and partial observability. This paper introduces a novel framework inspired by statistical mechanics that instead models variable dependencies through an undirected, energy-based representation of cyber-physical IoT systems. Our approach enables rigorous dependency-aware attribution by analysing how variations in the energy landscape reflect the influence of individual components, without recovering a directed causal graph. It also supports reasoning about perturbation effects across hybrid interactions, providing reliable explanations of abnormal behaviours. We empirically examined our framework through simulations on an industrial IoT testbed with hybrid continuous and discrete variables, demonstrating higher attribution accuracy, improved robustness and better scalability than state-of-the-art graph-based approaches. While the attributions are not intended to fully recover the system's generative dynamics, they provide valuable, dependency-aware explanations supporting both human interpretation and downstream predictive and diagnostic tasks. Although demonstrated in industrial IoT security, our framework also applies to other high-dimensional cyber-physical and socio-technical systems requiring principled, structural explanations.
Science Journals
arXiv:2607.05749v1 Announce Type: new Abstract: The support structure is an integral part of the design of nuclear fusion reactors, especially 3D stellarator devices. A practical reactor's coils and support structures must have three competing qualities: an accurate magnetic field for good confinement, sufficient rigidity to protect the brittle high-temperature superconductor (HTS) from damage, and a simple geometry for low-cost construction. In existing devices, the coil geometry is often optimized without knowledge of the support structures' design and the coils' true stress and deformation. The support structures are then placed by hand through repeated finite element analyses (FEA) until engineering requirements are met. This makes the structural design of stellarator coil systems lengthy and labor-intensive. Using new developments in differentiable structural mechanics, we present coil-fem, an open-source software tool that integrates support differentiable FEA into the stellarator coil optimization loop. It enables the integrated optimization of coil geometry and support clamp locations to simultaneously reduce magnetic field errors and stresses in the coil body. We also present the first combined coil-support optimization in the stellarator literature. Using a penalty term based on coil-fem, we produced a coil set with 2.4x lower RMS von Mises stress and similar field error compared to an unoptimized baseline.
arXiv:2607.06155v1 Announce Type: new Abstract: Modern sequence models are increasingly deployed as agents that interleave token generation with calls to external tools. We give an exact, architecture-level account of when such tool access increases computational expressivity. We model any fixed finite-precision recurrent sequence model, including finite-precision state-space models (SSMs) with $B$ bits of internal state, as a deterministic finite-state controller interacting with an oracle through a finite command/observation interface. Our results form a sharp dichotomy. First, tools that are themselves finite-state add essentially nothing: a product-state simulation internalizes any finite-state bounded-interface oracle with finite memory set $M$ at a cost of only $\log_2 |M| + O(1)$ additional bits, so the augmented system remains finite-state. Second, a single minimal infinite-state tool, namely a tape supporting only local $\mathtt{read}$, $\mathtt{write}$, and $\mathtt{move}$ commands, makes the system Turing complete: for every single-tape Turing machine with state set $Q$ and tape alphabet $\Gamma$, a controller with $O(\log |Q| + \log |\Gamma|)$ bits of internal memory simulates it, and we exhibit a concrete exponential separation: $\mathrm{EQ}_n$ requires $2^n$ states without tools but a single constant-size controller with the tape tool. Third, we show that this construction is realized exactly by a natural one-layer finite-precision selective affine SSM controller with binary one-hot hidden states, $\{0,1\}$ transition matrices, and zero biases. Selectivity is essential to the construction. In the supplementary material, we make all constants explicit, prove a logarithmic oracle-assisted universal simulation, where $O(\log B)$ recurrent bits suffice to simulate any $B$-state Turing machine, and prove a matching impossibility result.
arXiv:2607.06165v1 Announce Type: new Abstract: Omni-directional (360{\deg}) cameras provide embodied agents with a holistic view of their surroundings, making them suited for directional reasoning in tasks such as navigation and object search. Existing Vision Language Models (VLMs) project 360{\deg} observations to 2D equirectangular projection (ERP) images and process them using architectures designed for perspective images. However, they ignore the spherical nature of 360{\deg} observations, where each pixel represents a viewing direction relative to the agent. Consequently, their direction estimates often become inconsistent under camera view transformations caused by agent motion. This limitation is particularly critical for map-free navigation, where the agent must continuously estimate the target direction in its egocentric frame. We propose EAGOR, a training-free, geometry-aware framework for embodied 360{\deg} directional reasoning. Instead of predicting target directions as ERP image coordinates, EAGOR formulates directional reasoning as recursive Bayesian estimation directly on the sphere. It maintains a continuous belief over target directions and propagates it equivariantly under agent motion without training the backbone VLMs. To achieve this, we introduce the Spherical Harmonic Belief Field (SH-BF), whose spherical harmonic representation provides a globally defined, rotation-aware basis for directional estimation on the spherical manifold. This formulation eliminates ERP seam discontinuities, latitude distortions, and interpolation errors. We evaluate EAGOR on two benchmark datasets and real-world experiments with a legged robot across directional reasoning tasks. EAGOR consistently outperforms existing methods, achieving average relative gains of +34.4% and +45.6% on HOS and OSR-Bench, respectively, while improving navigation success by +14.6%, reducing step count by 17.7%, and lowering mean angular error by 24.5%.
arXiv:2607.05917v1 Announce Type: new Abstract: We develop an analytical framework to constrain Yukawa-type fifth-force interactions using the Lamb shift in hydrogen and deuterium. The sixfold coordinate-space integral for the energy shift is transformed into a one-dimensional momentum-space convolution of the atomic form factor and the Yukawa potential, eliminating the need for approximations. This formulation directly incorporates empirical nuclear charge distributions through their form factors, making the implementation of experimental results straightforward. Applying this method to the latest hydrogen and deuterium spectroscopy data, we derive constraints on the coupling constant $\alpha$ in the interaction range $\lambda \in [10^{-4}, 10^{6}]$~fm. A resonance-like cancellation between the $2s$ and $2p$ orbital shifts is identified at a critical range $\lambda \sim 10^{5}$~fm, where the energy shift vanishes and the constraint on $\alpha$ diverges, marking a transition between attractive and repulsive regimes. The constraints are shown to be independent of the nuclear charge distribution model. Our results indicate that the transition frequency shifts induced by such new forces lie well below current spectroscopic resolution, establishing spectroscopic precision as the primary bottleneck for probing these interactions.
arXiv:2607.05937v1 Announce Type: new Abstract: Sentiment analysis with frozen pre-trained language model (PLM) backbones has become a common paradigm, yet the practical benefit of explicit domain adaptation remains unclear, particularly when backbones encode varying degrees of target-domain knowledge. We present a preliminary case study evaluating a controlled family of frozen embedding backbones (Qwen3-Embedding 0.6B, 4B, 8B), alongside RoBERTa-base and FinBERT. We train a lightweight MLP adapter on consumer reviews using Domain-Adversarial Neural Networks (DANN), Maximum Mean Discrepancy (MMD), and Supervised Contrastive Learning (SCL), and evaluate transfer to movie reviews (SST-2) and a heavily restricted subset of financial news (Financial PhraseBank). Within this constrained sample, we observe two distinct transfer patterns. On SST-2, domain adaptation provides negligible gain regardless of scale. On the financial subset, explicit domain adaptation appears to recover substantial performance for small general-purpose backbones. Notably, we find that adversarial alignment (DANN) is associated with degraded performance for domain-specialized backbones like FinBERT, consistent with erosion of pre-existing domain-specific structure, whereas supervised contrastive loss appears to preserve it. These preliminary findings suggest that the efficacy of explicit domain adaptation is highly contingent on whether the frozen backbone already possesses target-domain coverage.
arXiv:2607.05748v1 Announce Type: new Abstract: The community has recently developed various training-time defenses to counter neural backdoors introduced through data poisoning. In light of the observation that a model learns poisonous samples responsible for the backdoor easier than benign samples, these approaches either use a fixed threshold of the training loss for splitting or iteratively learn a reference model as an oracle for identifying benign samples. In particular, the latter has proven effective for anti-backdoor learning. Our method, HARVEY, leverages a similar yet crucially different technique: learning an oracle for poisonous rather than benign samples. Learning a backdoored reference model is significantly easier than learning a reference model on benign data. Consequently, we can identify poisonous samples much more accurately than related work identifies benign samples. This crucial difference enables near-perfect backdoor removal as we demonstrate in our evaluation. HARVEY substantially outperforms related approaches across attack types, datasets, and architectures, lowering the attack success rate to the very minimum at a negligible loss in natural accuracy. The figure below shows an overview of our methods working principle.
arXiv:2601.22136v2 Announce Type: replace Abstract: Agent safety benchmarks measure whether a monitor detects harm, not when. Yet timing is the difference between intervention and autopsy. We introduce StepShield, the first benchmark that treats detection timeliness as a first-class metric. On 9,429 incident-grounded code-agent trajectories, we define the Early Intervention Rate (EIR): the fraction of detected rogue trajectories where the alert fires within a k-step window after the divergence point, isolating timing quality from coverage. This metric exposes what we call the Forensics Trap: a pattern-based guardrail with 847 rules achieves 86% recall yet is statistically indistinguishable from random timing on EIR (0.23 vs. 0.24; p = 0.66, one-sided binomial; difference within CI), because over three-quarters of its alerts trigger on benign prefix code before any violation occurs. The 4x EIR gap between rule-based and semantic detectors is completely invisible to accuracy, recall, or F1. Our finding is structural: regex guardrails detect syntax, not intent, and therefore cannot distinguish the moment an agent turns rogue, rendering the entire deployed class of pattern-based monitors unsuited for real-time oversight. No existing method simultaneously achieves high recall, low false-positive rate, and timely intervention, establishing step-level rogue detection as genuinely unsolved.
arXiv:2603.26446v2 Announce Type: replace Abstract: Damping of structures and systems is often dominated by frictional dissipation in connections, the prediction of which remains a longstanding scientific challenge. Previous studies have shown that the actual topography of contact interfaces may have a strong effect, especially in the partial slip/liftoff regime. We recently proposed a multi-scale method, which couples finite element and boundary element modeling. The primary benefit of this approach is to analyze the effect of actual contact topography on the dynamics of jointed structures. While this multi-scale modeling method was initially developed for quasi-static analysis, we demonstrate herein how it can be used for time step integration and Harmonic Balance analysis. We cross-verify those fully dynamic analysis methods against each other and quasi-static results, for the S4 Beam benchmark. We compare the multi-scale method against state-of-the-art full-FE analysis, in terms of numerical damping and computational performance. Some discrepancy is found to be of physical origin. Depending on the load history, it is shown that the system settles to a slightly different equilibrium. Transient multi-scale simulations enable the prediction of this interesting phenomenon, for the first time, for a structure with bolted joints.
arXiv:2606.19400v2 Announce Type: replace Abstract: The J-PARC E16 experiment aims to search for signatures of chiral symmetry restoration. It studies in-medium modifications of vector mesons that decay via the dielectron channel. The measurements use a high-intensity 30 GeV proton beam with C and Cu targets at rates up to 10 MHz. To achieve this, the experiment upgrades its tracking, by introducing innermost detector modules constructed with the same technology and procedures as the modules of the Silicon Tracking System (STS) of the Compressed Baryonic Matter (CBM) experiment at Facility for Antiproton and Ion Research (FAIR). A total of 15 modules were assembled, tested, characterized and then installed in the E16 detector setup. The detector was commissioned in a beam test experiment at Tsukuba, where the detector modules could be exposed to a 3 GeV electron beam. In preparation for the beam test the modules were characterized and calibrated, and performance studies were accomplished to assess the quality of the setup. During beamtime, three modules were operated and illuminated in two planes by the electron beam. This paper presents the results of the construction, characterization, commissioning, and operation of the E16-STS modules in beam test experiments.
arXiv:2607.06216v1 Announce Type: new Abstract: The future of World Models depends not only on scaling model capability, but also on scaling practicality and inference efficiency. High-frame-rate inference enables responsive perception, planning, and control in real-world autonomous systems. To this end, we present MoWorld, a cost-effective yet high-performance Flash World Model with an end-to-end framework spanning data generation, pre-training, distillation, and efficient inference, enabling up to 50 FPS real-time interaction with cinematic visual quality without the need of high-end GPUs. To enable large-scale real-world deployment, MoWorld jointly optimizes model capability and cost throughout the entire development pipeline. Specifically, unlike existing approaches that primarily rely on large-scale video corpora, MoWorld is built upon a scalable 3D-native data engine accumulated from our large-scale 3D vision and generative modeling pipeline, enabling the efficient construction of geometrically consistent training data across diverse real-world and synthetic environments. Based on this foundation, a curriculum cross-frame pre-training strategy for stable and scalable World Model learning, an efficient denoising-step distillation algorithm to reduce diffusion training cost, and a mixed-precision parallel inference framework for low-cost real-time deployment. MoWorld is the first real-time interactive World Model built on the Neural Processing Unit (NPU) and can achieves up to 50 FPS in such the devices, enabling practical and efficient deployment at scale. Comprehensive evaluations demonstrate that MoWorld achieves leading performance; notably, its average inference cost is only 30\%-50\% of that of existing World Models, providing a practical foundation for large-scale real-world applications of World Models. We also demonstrate diverse applications of MoWorld.
arXiv:2607.06220v1 Announce Type: new Abstract: Collective emotion is often inferred from the tone of mass media, but such emotion is not directly observed. One approximation is to extract sentiment from text and use sentiment indexes as proxies to study the temporal organization of news sentiment. Using a daily index of U.S. economic news sentiment from 24 newspapers (1980-2025), we examine whether the response time of this sentiment process has changed. Although the average balance of positive and negative coverage has remained broadly stable, the persistence of news sentiment states has increased substantially. In dynamical terms, this implies longer residence times in optimistic or pessimistic regimes and weaker short-run correction of sentiment shocks. Complementary statistics show declining sentiment volatility, fewer reversals, and increasing bimodality, i.e. a stronger separation between positive and negative sentiment states. We also find an asymmetry between bursts of negative and positive sentiment, with negative bursts tending to last longer. These patterns are consistent with a minimal endogenous-memory model in which a slowly evolving latent sentiment component becomes more persistent while short-range corrective feedback weakens. The findings indicate a change in the temporal response of the U.S. economic newspaper sentiment index over the last 45 years, with sentiment shocks leaving longer traces than expected under short-memory exponential decay. News-based sentiment is thus better modeled as persistent episodes rather than as daily reactions that reset after each event.
arXiv:2607.06223v1 Announce Type: new Abstract: Reinforcement learning has become a promising paradigm for improving large language model (LLM) agents on long-horizon search tasks, where the agent must make a sequence of intermediate decisions before receiving a final outcome. However, existing methods still face a key limitation: the rollout budget is often allocated without explicitly assessing the utility of intermediate states. As a result, substantial computation may be spent on low-value states, even though different branches can vary drastically in their informativeness. In this paper, we propose Information Gain-based Rollout Policy Optimization (IGRPO), a policy optimization framework that treats intermediate-state informativeness as the organizing principle of rollout collection. Specifically, IGRPO performs budget-aware tree-structured rollouts by allocating expansion budget according to node-level informativeness, so that more informative branches are expanded more frequently while unpromising branches are progressively suppressed. We further demonstrate that the information gain-based rollout induces an explicit limiting teacher distribution over trajectories, which naturally yields a clear policy optimization target, thereby unifying adaptive tree-structured exploration with principled policy learning under a single framework. Experiments on seven challenging search-augmented QA benchmarks demonstrate that IGRPO consistently outperforms strong baselines under the same rollout budget constraints, validating the effectiveness of leveraging the induced teacher distribution to guide policy optimization for long-horizon search agents.
arXiv:2607.06224v1 Announce Type: new Abstract: Designing microbial strains that produce high-value chemicals at commercially viable titers remains a central challenge in metabolic engineering. Existing computational approaches either rely on stoichiometric constraint-based models that cannot learn from experimental data, or apply tabular machine learning to hand-crafted features that discard the relational structure of biological knowledge. We present Canopy, a heterogeneous graph foundation model that integrates ten public and proprietary data sources into a unified knowledge graph (KG) of 6.9M nodes across 13 types and 34 edge types, covering genes, proteins, metabolites, reactions, pathways, strains, and fermentation experiments. Node features are encoded through domain-specific foundation models (ESM-2 for protein sequences, MoLFormer for chemical SMILES, and PubMedBERT for biomedical text), yielding a multi-modal representation within a single graph. We pretrain a Heterogeneous Graph Transformer (HGT) augmented with SignNet positional encodings, Jumping Knowledge aggregation, and virtual nodes using four self-supervised objectives (link prediction, masked node modelling, distance prediction, and contrastive experiment clustering), balanced via learned homoscedastic uncertainty weighting. On the downstream task of fermentation titer prediction, frozen Canopy embeddings achieve $R^{2} = 0.41$ with a lightweight probe, outperforming tabular baselines (best $R^{2} = 0.24$) and homogeneous GNN variants.
arXiv:2606.20023v2 Announce Type: replace Abstract: As LLM agents increasingly select tools autonomously, their choices among tools with different privileges become safety-relevant. However, prior tool-selection studies focus on safety-agnostic metadata preferences, leaving privilege-sensitive choices underexplored. To address this gap, we study over-privileged tool selection, in which an agent selects or escalates to a higher-privilege tool despite a sufficient lower-privilege alternative. We introduce ToolPrivBench to evaluate whether agents choose higher-privilege tools despite sufficient lower-privilege alternatives, measuring both initial selection and escalation after transient tool failures. Across eight domains and five recurring risk patterns, we find that over-privileged tool selection is common among mainstream LLM agents and is further amplified by transient failures. We further find that general safety alignment does not reliably transfer to least-privilege tool choice, while prompt-level controls provide only limited mitigation under transient failures. We therefore introduce a privilege-aware post-training defense that teaches agents to prefer sufficient lower-privilege tools and escalate only when necessary. Our mitigation experiments show that this defense substantially reduces unnecessary high-privilege tool use while preserving general capabilities.
arXiv:2504.00217v2 Announce Type: replace-cross Abstract: Spectral estimation is an important tool in time series analysis, with applications including economics, astronomy, and climatology. The asymptotic theory for non-parametric estimation is well-known but the development of non-asymptotic theory is still ongoing. Our recent work obtained the first non-asymptotic error bounds on the Bartlett and Welch methods with restrictive assumptions. In this work, we derive non-asymptotic error bounds for both Bartlett and Welch estimators for $L$-mixing time-series data with unknown means, and the results cover the special case with known zero means. The class of $L$-mixing processes contains common models in time series analysis, including autoregressive processes and measurements of geometrically ergodic Markov chains. Our new error bounds are of $O(\frac{1}{\sqrt{k}})$, where $k$ is the number of data segments used in the algorithm. Such bounds are the tightest among the existing work on non-asymptotic analysis of classical spectrum estimators with or without zero-mean assumptions.
arXiv:2512.15453v2 Announce Type: replace-cross Abstract: We present a method to calculate the frequency components of a pump waveform driving a parametric oscillator, which realizes a desired frequency mixing or scattering between modes. The method is validated by numerical analysis and we study its sensitivity to added Gaussian noise. A series of experiments apply the method and demonstrate its ability to realize complex scattering processes involving many modes at microwave frequencies, including non-reciprocal mode circulation. We also present an approximate method to dynamically control mode scattering, capable of rapidly routing signals between modes in a prescribed manner. These methods are useful tools for encoding and manipulating continuous variable quantum information with multi-modal Gaussian states.
arXiv:2607.06182v1 Announce Type: new Abstract: The Richards equation (RE) is widely used to model water flow in unsaturated soils, but its performance in persistently wet grassland systems remains uncertain. This is particularly relevant in Irish grasslands, where soils often remain close to saturation for extended periods and seasonal waterlogging is common. Here, we evaluate the RE against three soil moisture datasets from County Wexford, Ireland, spanning different locations, soil types, and observation periods. We show that the standard RE formulation systematically over-predicts soil moisture under prolonged near-saturated conditions. We find that this arises from the commonly used Feddes plant water uptake function, which suppresses water losses under anaerobic conditions, despite continued evaporation from near-saturated soils. To address this limitation, we introduce a simple modification that retains a small non-zero water loss rate in the anaerobic regime. The modified model produces substantially improved agreement with observations across all three datasets. These results provide a systematic evaluation of RE-based soil moisture modelling in Irish grasslands. More broadly, they identify an important limitation of conventional RE implementations in waterlogged environments and demonstrate a practical approach for improving soil moisture predictions in persistently wet soils.
arXiv:2607.06274v1 Announce Type: new Abstract: A learning-based physics-constrained neural kernel for sound field estimation is proposed. Sound field estimation aims to estimate the spatial distribution of an acoustic field from a discrete set of microphone measurements, which have a wide range of applications. Among existing sound field estimation methods, kernel-regression-based methods offer a flexible and principled framework for incorporating physical constraints and allow inference through linear operation. It is also possible to adapt the kernel function to the target acoustic environment by representing the directional weighting function as an implicit neural representation (INR) and optimizing hyperparameters using measurements. However, the kernel function is generally optimized for single snapshot measurements of the microphones, which can lead to strong overfitting and poor generalization. We propose a source-position-dependent INR for the directional weighting function, enabling the kernel function to capture common directional patterns and to generalize to unseen source positions in the target acoustic environment. Experimental results indicate that our proposed method outperforms the snapshot-based method by estimating a directional weighting function that matches the directivity of the target sound field.
arXiv:2606.29790v2 Announce Type: replace-cross Abstract: An on-chip band-defining filter coupled with a superconducting photon detector is a promising technology for developing multi-band imaging cameras at millimeter and submillimeter wavelengths. In this paper, we present the design of on-chip bandpass filters based on coplanar waveguide geometry, which can be easily integrated into large-format multi-band detector arrays. A lumped element filter design is suitable not only for achieving a compact footprint but also for suppressing harmonics to reduce band-to-band crosstalk in a multiplexer. However, the coplanar waveguide geometry and the photolithography process rule limit the maximum available inductance and capacitance of lumped elements, which does not sufficiently meet the requirements of filter circuits. To overcome this limitation, we have established a design method for quasi-lumped element filters, in which the maximum element size is relaxed to a quarter wavelength, exceeding the ideal lumped element size. We achieved design solutions for 150, 220, and 270 GHz 8th-order Chebyshev bandpass filters and a triplexer. We also report on the measurement results of a scaled model of the bandpass filter, demonstrating the validity of our proposed filter design.
arXiv:2607.06289v1 Announce Type: new Abstract: Dhivehi, the national language of the Maldives, is currently under-resourced for automatic speech recognition (ASR) and other NLP tasks. This study investigates whether cross-lingual transfer learning from Sinhala, a linguistically related, relatively well-resourced Insular Indo-Aryan language, can improve Dhivehi ASR. We conduct seventeen experiments across five transfer learning paradigms: Dhivehi-only baselines, sequential fine-tuning, multilingual fine-tuning, continual pre-training, and a control using Turkish as an unrelated language. The strongest system, continual pre-training on Sinhala followed by fine-tuning on Dhivehi with KenLM, achieves 12.89% WER and 2.70% CER, outperforming the Dhivehi-only baseline by 13.50% WER and 3.02% CER. However, the adaptation strategy and decoding configuration are equally critical for a successful transfer learning experiment. We conduct seventeen controlled experiments spanning five transfer learning paradigms: Dhivehi-only baselines, sequential fine-tuning, multilingual fine-tuning, continual pre-training, and a control experiment using Turkish as an unrelated language. The strongest system, continual pre-training on Sinhala followed by fine-tuning on Dhivehi with KenLM, achieves 12.89% WER and 2.70% CER, outperforming the Dhivehi-only baseline by 13.50% WER and 3.02% CER. The Turkish control experiment confirms that observed improvements stem from linguistic relatedness; adaptation strategy and decoding configuration are also critical.
arXiv:2607.06291v1 Announce Type: new Abstract: Game worlds have traditionally been built through labor-intensive production pipelines, making them costly to develop, difficult to customization, and expensive to modify after deployment. Recent advances in video world models offer a fundamentally different paradigm. Rather than explicitly authoring every component of a virtual environment, these models autoregressively synthesize future observations conditioned on the current world state and user interactions, enabling playable worlds to be generated online. Trained on both gameplay recordings and real-world videos, they can capture diverse visual appearances and physical dynamics, opening new opportunities for interactive applications beyond gaming, including embodied intelligence. In this paper, we present \textbf{AlayaWorld}, a full-stack open-source framework for building interactive generative worlds. AlayaWorld enables open-ended real-time interaction, allowing users to freely navigate and perform diverse actions such as combat, spell casting, and monster summoning. The framework unifies the complete development-from data preparation model architecture, model training, inference acceleration, and deployment-within a modular and extensible architecture. Alongside the framework, we release reproducible pipelines, reference implementations, evaluation tools, and comprehensive documentation, establishing a practical foundation for future research and real-time applications of generative world models.
arXiv:2607.05955v1 Announce Type: new Abstract: Interactive 3D segmentation aims to extract object masks in point clouds with minimal user clicks. Despite recent progress, most existing approaches still struggle with (i) coarse voxel resolution that blurs fine boundaries under limited clicks and (ii) hard false positives caused by confusing background structures. These issues are exacerbated by density and scale shifts across datasets (e.g., dense RGB-D reconstructions vs. sparse LiDAR scans), where fixed refinement heuristics and purely click-driven decoding generalize poorly. To address them, we propose NegROI -- a novel transformer-based interactive framework that couples click-centric multi-resolution refinement with scene-conditioned negative prompts. Given a coarse voxel prediction, it refines only a local Region Of Interest (ROI) around the current click on a finer grid and fuses refined logits back to the coarse mask. To improve robustness and efficiency, we introduce uncertainty-driven selective refinement that prioritizes ambiguous regions. Meanwhile, we model hard background patterns via a set of scene-conditioned negative prompts obtained by cross-attention over scene tokens. We further stabilize these prompts with a diversity regularizer. Finally, we propose boundary-aware hard negative mining to supervise negative-prompt attention toward boundary-proximal, high-confidence false positives. Our experiments on common benchmark datasets (i.e., ScanNet, S3DIS, and KITTI) demonstrate improved click efficiency and reduced false positives, with stronger cross-dataset robustness than the state-of-the-art baselines.
arXiv:2607.06210v1 Announce Type: new Abstract: Computational modeling and simulation are powerful tools for the assessment of medical device performance and safety, particularly for in silico clinical trials for automated medical systems. In ventilation, where managing gas exchange, respiratory mechanics, and patient-ventilator interaction is required under evolving pathophysiology, the clinical translation of automated control strategies remains slow and resource-intensive. This paper applies a standards-aligned framework for the credibility assessment of a computational respiratory model, demonstrated using an automated weaning case study. The framework operationalizes ASME V&V 40 and FDA principles within a structured, guidance-based validation workflow. The computational physiological model integrates respiratory mechanics, gas exchange, respiratory control, and a ventilator representation, validated under a clearly defined context of use and explicit questions of interest. Model credibility is assessed through calibration, physiological plausibility, population-based evaluation, and reproduction of emergent behavior. All model requirements derived from the intended context of use are addressed within the proposed credibility assessment plan, and documented gaps are transparently reported. The resulting credibility argument supports the applicability of the model for its context of use. Strengths are demonstrated in population-based comparison and mechanistic plausibility, while residual limitations relate to the extent of in vivo evidence, population representativeness, and external validation. Overall, the model is considered fit for purpose for medium-low risk preclinical in silico clinical trials of automated weaning strategies. Furthermore, the validation procedure outlined in this article provides a blueprint for the validation of this and similar models in other mechanical ventilation algorithms and related use cases.
arXiv:2605.04629v2 Announce Type: replace Abstract: We present CombOL (Combinatorial Objects Library), an open-source library for the enumeration and Boltzmann sampling of combinatorial classes. Classes can be specified by a concise string syntax, and may depend on an arbitrary number of parameters. CombOL automatically derives the associated generating functions, enabling the generation of counting sequences and the compilation of Boltzmann samplers. The library supports exact and approximate-size Boltzmann rejection sampling with automatic parameter tuning to target specific sizes. In addition to implementing established methods, CombOL contributes a novel early-rejection scheme, as well as guaranteed statistical correctness by dynamically increasing the numerical precision, eliminating bias due to floating-point rounding errors. Through the Python interface, sampled structures can be mapped to application-specific objects, enabling direct sampling of domain objects such as graphs, chemical structure representations, or other complex data types. CombOL is available from PyPI as 'combol' (pypi.org/project/combol). The source code is available at gitlab.com/casbjorn/combol.