arXiv:2606.27059v1 Announce Type: new Abstract: We develop a type system for secure information flow where new security levels can be created and inserted into the security lattice dynamically, i.e., even in the middle of an execution of a system. Our system is formalized by extending Kobayashi's type-based secure information flow analysis for Milner's pi-calculus, which is one of the most expressive models (or "languages") supporting both sequential and concurrent computations, with concise syntax, reduction-based semantics, and bisimulation equivalence as a robust formalization of secrecy as non-interference. The development required careful treatment of extensions of lattices themselves as well as deliberate generalization from the simple 2-element lattice (consisting of only High and Low) in the original system.
Science Journals
arXiv:2606.26180v1 Announce Type: new Abstract: Modern software systems routinely need the same data model in several schema languages: a model may exist as JSON Schema for a web API, as XSD for data exchange, and as SHACL for a knowledge graph. Keeping these representations consistent as the model evolves is a recurring construction and maintenance burden, because converters between schema languages are hard to find, scattered across ecosystems, of uneven quality, and frequently lossy. We study, empirically, to what extent such imperfect, heterogeneous converters can be orchestrated into automated, reproducible, and quality-ranked conversions, and where the current converter landscape reaches its limits. Our approach models schema languages as nodes and converters, treated as black boxes, as directed edges, so that conversions become paths that are discovered, executed, ranked, and reported with full per-step provenance, with failures handled by trying alternatives. We realize it as the open-source Schema Conversion Orchestrator, integrate it into MetaConfigurator, and evaluate it on 60 conversion tasks built from real-world schemas across five schema languages, using agent-assisted, human-reviewed quality annotations. Orchestration surfaces a usable result for 43 of 60 tasks; the remaining failures localize concrete gaps in the converter landscape. We discuss implications for tool builders and for measuring conversion quality.
arXiv:2606.26185v1 Announce Type: new Abstract: LLM-as-judge ("grader") components are now standard in evaluation harnesses, including safety evaluations where a pass/fail verdict may gate downstream deployment decisions. A widespread assumption is that setting the grader's sampling temperature to 0 makes grading deterministic. We test this assumption against a real safety-evaluation codebase (Japan AISI's open-source aisev) and show it fails on two levels. First, the harness invokes its grader without setting temperature or seed; the underlying provider silently applies its default of 1.0, so items near the decision boundary flip pass/fail across identical runs (per-item disagreement up to ~50% over 20 runs). Second, pinning temperature=0 reduces but does not eliminate flips: across 690 API calls spanning two providers, three model tiers, and five sampling configurations, 1-2 of 7 borderline items remain non-reproducible even under forced greedy decoding (top_k=1). Claude Opus 4.7/4.8 has since deprecated temperature entirely, rendering the primary mitigation inapplicable to newer model generations. These findings expose a structural gap: evaluation harnesses that report single-run verdicts without variance or grader-disagreement metrics can present noise as a safety property. We release a reproduction harness (690 calls, 7 conditions) and recommend that harnesses treat grader disagreement as a first-class health metric alongside the scores themselves.
arXiv:2606.26244v1 Announce Type: new Abstract: We introduce an approach to distinguishing isomorphism types of graphs based on vector spaces of polynomials that are set-wise invariant under permutations ("separating modules," which are representations of the symmetric group), inspired by the Geometric Complexity Theory approach to separating complexity classes (Mulmuley & Sohoni, SIAM J. Comput., 2001). We characterize the power of this method for distinguishing non-isomorphic graphs under several different complexity measures: - We show that separating modules of "support-degree" $k$ (each monomial touches at most $k$ vertices) are equivalent to the counts of $O(k)$-vertex subgraphs. This is strictly weaker than $O(k)$-dimensional Weisfeiler--Leman (F\"urer, ICALP '01). - We show that separating modules of symmetric circuit size $n^{\Theta(k)}$ are equivalent to $\Theta(k)$-WL. This generalizes and strengthens a result of Dawar & Wilsenach (CSL '18; ICALP '20; ACM Trans. Comput. Log., 2022; Theory Comput., 2025): they proved one direction of this equivalence for invariant polynomials; we generalize to separating modules and prove both directions. - When considering only the multiplicities of separating modules (as was proposed in GCT by Mulmuley & Sohoni, ibid., rather than the polynomials themselves), we show that two graphs are separated by multiplicities if and only if their automorphism groups have different cycle indices. The latter result is notable in the analogy with GCT, as it is the only result we are aware of in which the multiplicity approach to separating isomorphism types of objects has been given an "intrinsic" characterization in terms of the objects themselves. We use this to show that for graphs, multiplicity obstructions are stronger than occurrence obstructions. We also connect invariant polynomials to the Graph Reconstruction Conjectures and Forman's "invariants of finite type" (Adv. Math., 2004).
arXiv:2606.26699v1 Announce Type: new Abstract: Machine learning is emerging as a valuable tool for convection-related applications such as post-processing numerical weather prediction output, improving understanding of convective storm climatology and potentially improving existing convective parameterization schemes. In a rapidly developing field, it is vital to assess the strengths and limitations of machine learning approaches across different applications. Here, a probabilistic model is developed using a convective cell dataset as ground truth and predictors primarily from ERA5. The model's ability to reproduce the convective cell climatology at different regions relative to cold fronts (i.e. post-frontal and pre-frontal) is assessed during the warm-season in Germany. The optimal number of features (predictors) is selected using a feature elimination strategy. Overall, the optimised model exhibits high skill in reproducing the spatial and temporal cell frequency at different regions relative to the front. While the highest cell frequency is correctly identified near the surface front, the model underestimates the actual cell count in this region. Feature importance analysis shows that the model depends most heavily on CAPE to make its predictions. Additionally, the time of day predictor is key for accurately capturing the diurnal cycle of convective cells on both sides of the cold front. The study highlights both the advantages and the limitations of data-driven models, offering valuable insights for future data-driven climate and weather prediction models.
arXiv:2606.26808v1 Announce Type: new Abstract: This work investigates the impact of reconfigurable intelligent surfaces (RIS) on radio links other than the one for which the RIS configuration is optimized. We consider three different scenarios in which a secondary communication link could be affected by a RIS whose configuration is optimized for a primary communication link operating in the vicinity, on the same or on different frequencies. This question is investigated experimentally in the FR1 band, using the CorteXlab radio testbed and a Greenerwave RIS. We show that the impact, in terms of received power and impact on the channel phase of the secondary link, is significant even outside of the nominal frequency range of the RIS, and is not mitigated by carrier frequency separation between the two communication links.
arXiv:2606.26839v1 Announce Type: new Abstract: Learning robust navigation policies directly from visual observations remains a fundamental challenge in vision-based robotic navigation. In end-to-end imitation learning approaches, the visual encoder and action decoder are jointly optimized using a single action loss, which provides only an indirect supervisory signal to the encoder. This indirect supervision frequently results in the encoder learning ambiguous, action-agnostic representations. The problem is further complicated by substantial variations in scene structure and appearance across diverse environments, as well as the prevalence of visual distractors inherent to real-world navigation settings. Such action-agnostic features cause the navigation policy to produce inconsistent actions at ambiguous decision points, leading to navigation failure. To overcome these limitations, we propose ORION (Ordinal Neural Collapse for Visual Navigation), a method that explicitly organizes the encoder's representation space according to the ordinal structure of navigation actions. In the context of goal-directed navigation, ego-centric control categories from Far Left to Far Right exhibit a natural ordinal relationship in which neighboring classes share similar visual contexts, while semantically opposing classes differ substantially in appearance. We encourage class representations to be arranged sequentially along a single discriminative axis, while suppressing off-axis variance within each class. The pretrained encoder is then integrated into a diffusion-based navigation framework, and the full pipeline is fine-tuned end-to-end. Extensive experiments in both simulation and real-world settings show that ORION consistently outperforms end-to-end and neural collapse baselines in navigation success rate and goal progress, with notable gains in visually challenging scenarios such as complex multi-way intersections.
arXiv:2606.26858v1 Announce Type: new Abstract: Reinforcement learning (RL) has become a prominent framework for developing driving experts in autonomous vehicles. However, most existing RL-based experts are designed to output direct control commands (e.g., throttle, steering), which suffer from a lack of interpretability, high spatial complexity in learning road geometries, and poor compatibility with modern end-to-end planning architectures. To address these limitations, we propose a novel trajectory planning architecture for RL driving experts that integrates an RL policy with a polynomial-based trajectory planner. By employing a Frenet-frame coordinate system, our method simplifies complex road geometries into a curvilinear framework, offering a structured coordinate prior that facilitates policy learning. Furthermore, we incorporate a kinematic feasibility check into the planning stage to ensure that generated trajectories remain within the vehicle's physical limits, effectively mitigating cumulative tracking errors typically found in planning-based systems. We evaluate our approach on key CARLA benchmarks, where it significantly outperforms existing state-of-the-art control-based RL experts. On the CARLA Offline Leaderboard v1 and NoCrash benchmarks, our method improves the driving score by 5% and 11%, respectively, and increases the success rate by 8% and 19%.
arXiv:2606.26861v1 Announce Type: new Abstract: Deploying large language models (LLMs) on Industrial Internet of Things (IIoT) edge devices demands extreme compression, yet existing structured pruning methods collapse at high compression ratios due to one-shot importance estimation, and their cross-architecture behavior remains unpredictable. This article presents a cascaded multi-granularity pruning framework that removes layers, attention heads, and feed-forward channels in coarse-to-fine order, with lightweight low-rank recovery between stages to re-estimate component importance. An information-theoretic analysis motivates this ordering, and the Structural Independence Assumption (SIA) is formalized as a checkable condition predicting whether per-component pruning criteria are reliable for a given architecture: Multi-Head Attention (MHA)+GELU designs satisfy the SIA, whereas Grouped Query Attention (GQA)+SwiGLU designs violate it. On bearing fault diagnosis spanning 88M to 6.25B-parameter models, the framework extends achievable compression to 13.8 times on MHA+GELU architectures with 83.82% accuracy (+3.70 percentage points (pp) over the strongest baseline), while exposing a ~74pp accuracy collapse on GQA+SwiGLU architectures that violate the SIA. Deployed on an industrial slewing bearing fault diagnosis platform with NVIDIA DGX Spark, compressed models reduce inference latency by up to 67.2% and peak memory by 62.5%, demonstrating viability for IIoT edge inference.
arXiv:2606.26884v1 Announce Type: new Abstract: We present MedSWFlow, an open-source, model-agnostic LLM workflow for drafting medical social work case plans. The framework translates professional case-planning tasks into six stages: assessment, problem analysis, goal setting, intervention planning, risk anticipation, and planned effect evaluation. Drawing on established social work and behavioral frameworks, MedSWFlow standardizes case inputs, builds structured case profiles, and generates reviewable assessment forms and service plans through staged prompting. The system is released as an open-source research framework for reproducible case-plan generation across LLM providers. Outputs are intended as practitioner-reviewed drafts rather than final service decisions. Source code: https://github.com/santhiyacw-droid/MedSWFlow/tree/main.
arXiv:2606.26886v1 Announce Type: new Abstract: Artificial intelligence (AI) systems for automated Critical View of Safety (CVS) assessment in laparoscopic cholecystectomy are nearing clinical translation. Beyond algorithmic performance, clinical safety and effectiveness depend on the quality of the human-machine interface (HMI). This work examines how AI-generated predictions should be presented and controlled intraoperatively. Seventeen surgeons, including residents, attending surgeons, and professors, took part in a mixed-methods, user-centered design study to optimize an intraoperative HMI for AI-assisted safe laparoscopic cholecystectomy. Interviews explored interaction modalities, timing of assistance, visualization strategies, and control mechanisms across surgical roles, and were analyzed using reflexive thematic analysis and human-factors heuristics. Most surgeons (16/17) supported the use of AI for intraoperative decision support while rejecting autonomous decision-making. Attendings preferred minimal AI feedback at decisive moments (13/14), whereas residents favored optional guidance (3/3) with confidence indicators and on-demand anatomical overlays. Across interviews, surgeons consistently prioritized visual, surgeon-controlled, minimally intrusive displays, with the strongest support for a minimal overlay (16/17) and on-demand anatomical segmentation (13/17). Recurrent concerns included persistent overlays, haptic feedback, and numeric confidence displays, although these were not uniformly raised across the cohort. These findings informed the design of CVS Copilot, a surgeon-controlled, role-adaptive HMI that provides AI-based CVS assessment with minimal default visualization and optional overlays.
arXiv:2606.26891v1 Announce Type: new Abstract: Concept Bottleneck Models (CBMs) promise transparent reasoning by predicting through human-interpretable concepts, yet their effectiveness fundamentally depends on how well visual and textual representations are aligned or matched. Existing vision-language CBMs often rely on pre-aligned encoders or global cosine similarity, which obscures fine-grained concept localization and fails to reflect true semantic geometry. In this work, we rethink concept alignment as a dynamic cross-modal transport process instead of static projection and propose the Optimal Transport Flow Concept Bottleneck Model (OTF-CBM). It first learns a data-driven semantic cost via Inverse Optimal Transport to measure cross-modal distances, and then performs unbalanced optimal-transport-based flow matching to model semantic transitions between visual patches and textual concepts. With velocity-based concept activation, OTF-CBM captures interpretable geometric relations without ODE integration. Experiments further show that OTF-CBM achieves superior classification accuracy and concept faithfulness, offering a new geometric and dynamical perspective for interpretable cross-modal reasoning.
arXiv:2606.26894v1 Announce Type: new Abstract: Brain MRI poses a fundamental challenge for machine learning: models must learn from high-dimensional 3D data spanning multiple co-registered modalities, despite the limited sample sizes typical of neuroimaging studies relative to the diversity in anatomy, pathology, and acquisition conditions. While multimodal imaging provides complementary information critical for clinical interpretation, effectively integrating these signals remains difficult. We propose Multimodal Intra- and Cross-Context Vision Transformer (MICViT), a 3D vision transformer that explicitly models both modality-specific representations and cross-modal interactions across local and global contexts. Concretely, MICViT combines four attention mechanisms: modality-specific local and global attention for intra-modal feature learning, and cross-modal local and global attention to capture interactions between modalities. We evaluate MICViT on brain age prediction across three heterogeneous datasets (UK Biobank, n=41,404; SOOP, n=1,062; Cam-CAN, n=613) using multiple MRI modalities (e.g. T1, FLAIR, DWI, SWI). MICViT consistently outperforms state-of-the-art CNN and transformer baselines in 3D settings. Notably, it benefits more strongly from multimodal inputs, yielding larger performance gains as additional modalities are incorporated. These results demonstrate that explicitly modeling intra- and cross-modal interactions is key to unlocking the full potential of multimodal brain MRI, highlighting a promising direction for representation learning in neuroimaging.
arXiv:2606.26900v1 Announce Type: new Abstract: High-dimensional tensor data streams arise naturally in scientific and engineering applications, such as simulations of kinetic equations and quantum systems, where samples become available sequentially and are often already represented in compressed low-rank tensor formats. Existing streaming tensor-train (TT) algorithms typically construct or update representations from dense tensor data or randomized sketches. However, when high-dimensional data are generated directly in TT or related low-rank formats, reconstructing dense tensors solely for the purpose of compression is unnecessary and computationally prohibitive. We develop a deterministic incremental TT compression algorithm that operates directly on streaming TT-formatted data. Given a new TT tensor, the proposed method updates an accumulated TT representation through core-wise projection, residual orthogonalization, and adaptive enrichment, retaining only the complementary information that cannot be represented within a prescribed tolerance. By operating entirely at the level of TT cores, the algorithm avoids reconstructing either the incoming tensor or the accumulated full tensor. We establish approximation error bounds for the proposed incremental approach. Moreover, we show that the accumulated TT representation corresponds to a compressed analogue of standard proper orthogonal decomposition for full-order snapshot data, enabling reduced-order models to be constructed directly from streaming low-rank solution data through operations on TT cores, without first reconstructing full snapshots. Numerical experiments on parametric radiative transfer equations demonstrate that the proposed method achieves comparable reconstruction accuracy with substantially reduced wall time and yields efficient and accurate ROMs directly from compressed low-rank data.
arXiv:2606.26901v1 Announce Type: new Abstract: Automatic Speech Recognition (ASR) is increasingly used to document clinical encounters, yet its reliability in multilingual and demographically diverse Indian healthcare context remains largely unknown. In this study, we first conduct the systematic audit of ASR performance on real-world psychiatric interview data spanning Kannada, Hindi and Indian English, comparing eight state-of-the-art models including IndicWhisper, WhisperLargeV3, Sarvam, GoogleS2T, Gemma3n, OmniLingual, Vaani, and Gemini. Our results reveal substantial variability across models and languages, with some systems performing competitively in Indian English but failing in regional speech. We further fine-tune two of the best performing opensource models, i.e., Gemma3n and OmniLingual, using various methods. With this, we uncover systematic performance gaps tied to speaker role and gender, raising concerns about equitable deployment in clinical settings, which are further mitigated by fairness-aware fine-tuning. To this end, we propose SamaVaani, a unified debiasing technique that simultaneously improves ASR performance and improves fairness across demographic groups.
arXiv:2606.26927v1 Announce Type: new Abstract: Microservice systems are prone to recurrent architectural anti-patterns (APs) that hinder maintainability, evolvability, and operational quality. Most existing AP detection approaches rely on static analysis and handcrafted rules, which can be effective but are often tool-dependent, limited to explicitly encoded detection logic, and difficult to adapt to heterogeneous repositories. In this paper, we investigate whether large language models (LLMs) are ready to support architectural anti-pattern detection in microservice architectures through a prompt-based analysis pipeline over static repository artifacts. We evaluate three general-purpose LLMs on a curated benchmark of microservice repositories annotated with 16 architectural anti-patterns, and compare their performance against the state-of-the-art static-analysis tool MARS using a uniform evaluation protocol based on precision and recall. Our results show that LLMs can provide useful support for anti-pattern detection, achieving competitive performance on several anti-patterns, especially when the relevant evidence is local, heterogeneous, or semantically rich. At the same time, they exhibit clear limitations on anti-patterns that require explicit structural or cross-service dependency evidence, where static analysis remains more reliable. These findings suggest that LLMs are not yet a replacement for traditional analyzers, but already represent a promising complementary aid for architectural assessment in microservice systems.
arXiv:2606.26937v1 Announce Type: new Abstract: The engineering of adaptive user interfaces has traditionally relied on either rule-based systems encoding designer intuitions about user needs or machine learning approaches requiring substantial historical data before achieving effective personalization. We present a technical architecture that leverages Large Language Models as behavioral synthesis engines to enable immediate adaptation from sparse, heterogeneous user signals. Our system integrates three distinct behavioral channels, i) explicit micro-feedback on individual interface elements, ii) spatial priority inferred from manual widget reorganization through drag-and-drop interaction, iii) and attentional investment measured through dwell time during hover events, within a structured prompt engineering framework that continuously regenerates dashboard layouts while maintaining explanatory coherence. The architecture addresses the technical challenge of translating low-level interaction patterns into high-level design decisions through a layered prompt construction methodology that separates temporal context determination, behavioral signal extraction, explicit preference enforcement, and user profile synthesis. The approach combines manually specified behavioral interpretations and temporal heuristics with LLM-mediated synthesis, enabling the reconciliation of multiple simultaneous signals that would be difficult to encode through explicit rules alone. We demonstrate the system through an instantiation in the personal health monitoring domain, including an analytical evaluation of adaptation behavior across multiple scenarios and a working implementation managing fourteen distinct health metrics across seven widget visualization modalities. The evaluation compares profile-driven initialization, multi-signal behavioral adaptation, and presents the resulting interfaces through representative post-adaptation screenshots.
arXiv:2606.26943v1 Announce Type: new Abstract: Determining whether predicting two-dimensional sandpiles lies in $\mathbf{NC}$ or is $\mathbf{P}$-complete has been open for decades. Moore and Nilsson proved $\mathbf{P}$-completeness for the three dimensional case by encoding Boolean circuits into sandpiles, but this method fails in two dimension due to the impossibility of crossing gates. In this work, we study the existence of crossing gates on non-uniform and weighted grids. We establish an equivalence between uniform weighted crossing gates and a class of simple non-uniform crossing gates, which we call primal. We also exhibit a crossing gate that inherently requires more than one crossing, rather than a single crossing as in standard constructions. Finally, we show that the equivalence between uniform weighted and primal crossings breaks down in more general settings.
arXiv:2606.26947v1 Announce Type: new Abstract: While personalized image generation has achieved remarkable progress, multi-reference image generation (MRIG) remains a challenging task. Most existing benchmarks fail to adequately evaluate complex MRIG scenarios, hindering further progress in this area. To better assess model performance on complex MRIG tasks, we introduce OmniRef-Bench, a benchmark that covers complex combinations of reference image types and a large number of reference images. Evaluations on OmniRef-Bench show that mainstream open-source models struggle in complex MRIG scenarios, and their performance deteriorates significantly as the number of mixed-type reference images increases. To address this issue, we propose DyRef, a two-stage training framework. In the first stage, supervised fine-tuning equips the model with the basic capability to handle complex MRIG tasks. In the second stage, we introduce Difficulty-aware Advantage Reweighting (DAR) and Discriminative Reward Scaling (DRS). DAR dynamically adjusts the optimization objective to improve performance when handling a large number of mixed-type reference images. DRS enlarges intra-group reward differences for more effective policy optimization. Experiments demonstrate that DyRef significantly improves the performance of open-source models on OmniRef-Bench and single-image editing benchmarks, demonstrating the effectiveness and generalization capability of our approach.
arXiv:2606.26953v1 Announce Type: new Abstract: Premixed flames propagating in porous media or Hele-Shaw channels are governed by Darcy's law, which accounts for the strong frictional forces imposed by the solid matrix or confining walls. Prior theoretical studies of such flames have typically employed phenomenological Markstein-type corrections and have assumed unity Lewis numbers and adiabatic conditions. In this work, we develop a rigorous hydrodynamic theory for premixed flames under Darcy's law that incorporates nonunity Lewis numbers and heat losses. Using large activation-energy asymptotics and a systematic multiple-scale analysis, we derive the interfacial jump conditions across the flame from first principles. The conventional continuity requirements of mass flux and pressure at an interface under Darcy's law acquire corrections to the finite thickness of the flame. The adiabatic burning rate is shown to involve three distinct Markstein numbers, corresponding to curvature, tangential flow strain, and gravity-induced strain. The gravity term is unique to Darcy's law and has no counterpart in classical Navier--Stokes formulations. Moreover, the curvature Markstein number and the tangential strain Markstein number are found to be unequal, in contrast to the classical case where they coincide under constant transport properties. Explicit formulas for the Markstein numbers are provided, and the resulting new dispersion relation, linking the perturbation wave number $k$ to the growth rate $s$, takes the form $s = (a|k| - bk^2 - d|k|^3) / (1 + c|k|)$. This relation, applicable under Darcy's law, is to be compared to the classical Clavin--Garcia dispersion relation derived from the Navier--Stokes equations. The theory provides a rigorous foundation for flame dynamics in strongly confined environments, with direct applications to porous media combustion and Hele-Shaw cell experiments.
arXiv:2606.26967v1 Announce Type: new Abstract: Apple AirDrop and Google/Samsung Quick Share are proximity file-transfer protocols used by over five billion devices, yet their application-layer security properties remain largely unstudied because both stacks are proprietary and undocumented. Both protocols are reachable from wireless proximity without any prior pairing and process complex serialized content (binary plists, CPIO archives, Protocol Buffers, UKEY2 handshakes) inside privileged daemons, making them attractive zero-click targets across multiple operating systems. We perform the first cross-platform reverse engineering and protocol-aware fuzzing study of both stacks. We reconstruct AirDrop's seven-layer state machine and DVZip adaptive compression from binary analysis, build AIRFUZZ, a protocol-aware fuzzer that mutates pre-compression representations, and complement it with targeted hand-written analyses of Samsung's Quick Share service and Google's Quick Share for Windows. We discover six vulnerabilities (V1-V6): three pre-authentication issues in macOS/iOS AirDrop (V1: Swift fatalError DoS in the HTTP path router; V2: unbounded XML plist recursion in Foundation; V3: NULL dereference in Network.framework's HTTP/1.1 parser), two protocol-layer flaws in Samsung Quick Share (V4: pre-authentication OfflineFrame dispatch; V5: D2D encryption bypass for three frame types), and a heap use-after-free in Google Quick Share for Windows (V6) for which Google awarded a bounty. We responsibly disclosed all findings, and Apple, Samsung, and Google have acknowledged the reports.
arXiv:2606.26970v1 Announce Type: new Abstract: Introduction: Most Multiplayer Online Battle Arena (MOBA) analytics studies rely on structured data, which does not directly capture what each team could actually see during a match. Objective: This work introduces Dota2-Vis, a video-based dataset, and a baseline pipeline for visibility analysis in professional Dota 2 matches. Methodology: The dataset comprises all 144 matches from The International 2025, recorded from both team perspectives, totaling 288 Full HD videos, together with 2,477 manually annotated minimap images. We evaluate multiple variants of a modern object detector for player-icon detection and use the best-performing model to estimate opponent-visible player presence over time. Results: YOLO11l (large) achieved the best overall performance, reliably identifying player icons even in dense and visually cluttered minimap scenes. The resulting visibility curves reveal player, hero, role, and team-level patterns that complement conventional MOBA analytics, highlighting behavioral differences that are difficult to obtain from structured data alone. The dataset and code are publicly available at https://github.com/RicardoRCarvalho/dota2-vis/.
arXiv:2606.26976v1 Announce Type: new Abstract: Recently, Zhao & Dong (J. Fluid Mech. 2025, vol. 1013: A44) developed a high-efficiency, high-accuracy numerical framework, the shock-fitting harmonic linearised Navier-Stokes (SF-HLNS) approach, which enables a systematic study of the receptivity of non-modal perturbations in hypersonic blunt-body boundary layers over a wide parameter range. In this Part II, we employ a high-Reynolds-number asymptotic analysis to elucidate the physical mechanism of the receptivity process. A distinct slow-down convection mechanism is identified in the nose region, amplifying the perturbation streamwise vorticity from the post-shock position to the boundary layer around the stagnation point by a factor of O(\sqrt{R}), where R is the Reynolds number based on nose radius. Downstream, the lift-up mechanism further leads to a transient growth of the perturbation streamwise velocity up to an amplitude of O(R). Based on these mechanisms, a reduced model is developed to predict the downstream evolution of the non-modal perturbations initiated by receptivity, whose predictions agree well with SF-HLNS calculations. This model can also be used to investigate the effects of wall temperature and nose radius on non-modal receptivity efficiency, as will be detailed in Part III of this work series.
arXiv:2606.26986v1 Announce Type: new Abstract: Open Relation Extraction (OpenRE) requires a model to extract unseen relations between head and tail entities from unstructured text for real-world applications. The core challenge of OpenRE lies in achieving reliable generalization to unseen relation types. Current OpenRE approaches either employ clustering techniques, which cannot generate relation labels and suffer from poor generalization, or rely on direct relation label generation via Large Language Models (LLMs), which lack sufficient discriminative capacity to distinguish easily confused relations. To address these limitations, we propose Reasoning-guided progressive OpenRE (ReaORE), a framework for performing relation extraction through coarse-to-fine relation reasoning. Specifically, ReaORE consists of two key stages: (i) relation filtering, which reasons over multiple aspects to understand relations and instances, yielding an initial relation set, and further supplements and filters relations via embedding-based similarity to ensure the target relation is included; (ii) relation prediction, which aims to predict the target relations from the above set via fine-grained comparative reasoning to better distinguish easily confused relations. Extensive experiments on two widely used OpenRE datasets demonstrate that ReaORE outperforms existing baselines.
arXiv:2606.26987v1 Announce Type: new Abstract: Recent work identified emotion vectors in Claude Sonnet 4.5, which are internal representations that encode emotion concepts, causally influence behavior, and exhibit geometry mirroring human psychological structure. We test the generality of these findings in two open-weight models, Apertus-8B-Instruct-2509 and Gemma-4-E4B-it, extracting emotion contrast vectors across all layers, using two model-generated corpora. We recover valence geometry for both models, with peak PC1--valence correlations of $r = 0.76$ and $r = 0.83$, approaching the $r = 0.81$ reported for Claude.Beyond replication, we observe notable differences in how valence representations emerge across model depth. In Gemma-4-E4B-it, valence is strongly encoded in early layers but collapses towards later layers, whereas Apertus-8B-Instruct-2509 exhibits the opposite pattern, with valence representations absent in early layers, but emerging at mid depths. Arousal encoding, in contrast, is sensitive to the extraction corpus: both models show stronger PC2--arousal alignment with Gemma-generated stories ($r$ up to $0.45$) than Apertus-generated ones ($r \leq 0.21$), suggesting arousal-relevant cues are unevenly distributed across generated corpora. We open-source our experiment code and dataset for reproducible investigation of emotion representations across language model architectures.