Systems and Control
Papers
Fresh from arXiv, typeset properly. Highlighted words trade on the market.
Wednesday 12 August
Large Language Models (LLMs) are increasingly used in circuit design workflows, yet their reliability on simulator-facing SPICE netlist recognition and manipulation remains poorly understood and is rarely separated from high-level design reasoning. Although netlists are textual, they encode structured circuit objects through topology and parameters. We present NetlistBench, a structure-verified benchmark for SPICE netlist recognition and manipulation. NetlistBench contains 2,342 cases across 24 task families, covering parameter and connectivity recognition and edits, hierarchical operations, equivalence judgment, and long-horizon compound editing. Model outputs are evaluated by a deterministic structure-aware oracle. Across six non-thinking LLMs, performance varies substantially with operation-level structural complexity. Simple local edits reach $96\%$–$100\%$ accuracy, while device addition drops to $41\%$–$83\%$ and equivalence judgment to $49\%$–$90\%$. Enabling reasoning substantially improves weaker models but does not eliminate structure-preservation failures, with performance still degrading sharply as the edit horizon increases. NetlistBench identifies netlist reliability as a distinct bottleneck for trustworthy LLM-based circuit design automation.
Computer Vision
M-Net: Integrating Spectral Features and Physical Field Operators into Deep Learning for Medical Image Segmentation
Purpose: Deep learning-based medical image segmentation has achieved remarkable success, yet purely data-driven approaches often fail to exploit the rich mathematical structure inherent in medical images. We investigate whether explicit mathematical inductive biases, specifically matrix spectral analysis and vector calculus operators, can enhance segmentation beyond data-driven learning alone. Methods: We propose M-Net (Math-Augmented Network), which integrates three complementary mathematical priors into U-Net: (1) continuous spectral features derived from the condition number of centered local pixel matrices, providing a differentiable measure of texture ill-conditioning; (2) physical field operators (divergence and a discrete curl-like boundary irregularity operator) computed from image gradient fields, capturing focal intensity extrema and edge non-smoothness; and (3) a Math-Attention Gate (MAG) that adaptively fuses mathematical features with CNN-extracted deep features at skip connections. Results: Experiments on three benchmarks (LiTS, KiTS, and BraTS) show that M-Net achieves Dice scores of 78.42%, 76.15%, and 83.67%, outperforming baseline U-Net by 12.37%, 3.52%, and 5.55% on liver, kidney, and brain tumor segmentation, respectively. Ablations reveal that the condition-number feature contributes a 2.14% gain over binary invertibility features, while MAG adds 1.45% over simple concatenation. Conclusion: M-Net establishes that mathematical inductive biases provide effective complementary information for medical image segmentation. The continuous condition-number feature offers superior gradient information over discrete alternatives, and MAG preserves these priors throughout the network. This work opens avenues for integrating linear algebra and vector calculus into deep architectures for medical imaging.
Read the full abstract
Human-Computer Interaction
IF:CARGO: LLM-Based Semantic Compilation for Al-Native Rule Programming Games
This case study presents IF: CARGO, an experimental puzzle game that uses a large language model as a semantic compiler rather than an autonomous game-playing agent. Players author IF/THEN rules in natural language, which the model translates into a constrained command schema for deterministic validation and execution by the game engine. This architecture creates a playable loop of expression, execution, observation, and revision, framing AI interaction as semantic debugging. A mixed-methods playtest with 24 participants across eight levels examined player attempts, thinking time, perceived controllability, adjustability, and interpretations of the AI's role. Results suggest that players generally understood the model as a translation intermediary and could revise their strategies through feedback, while periodic commands, multi-robot coordination, and rule-priority mechanics created greater cognitive and diagnostic demands. The study proposes a practical pattern for AI-native gameplay: constrain natural-language input, preserve player authorship, and ensure deterministic execution.
Read the full abstract
Machine Learning
HYDRA: Hyperbolic Dynamic Representation Architecture for Kolmogorov-Arnold Networks
Kolmogorov-Arnold Networks (KANs) enhance nonlinear function approximation by replacing scalar weights with learnable univariate functions. However, assigning an independent function to every connection results in substantial parameter redundancy, limiting their scalability and efficiency. To reduce this redundancy, we introduce HYperbolic Dynamic Representation Architecture (HYDRA), a parameter-efficient hyperbolic extension of KAN that combines spline-based functional learning with representations in the Poincaré ball. HYDRA maps vector-valued inputs into a bounded hyperbolic latent space, performs KAN-style updates in tangent space, and employs a low-rank prototype block to share functional transformations across hidden dimensions. The resulting hyperbolic representations provide a structured radial coordinate for interpretation, while radius control improves training stability by preventing boundary saturation. Extensive experiments across eight benchmark datasets demonstrate that HYDRA consistently achieves competitive or superior predictive performance while improving parameter efficiency and representation interpretability.
Read the full abstract
Artificial Intelligence
How to Spend Your Oracle Budget: Practical Guidance for Protein Structure Prediction Models
Foundation models for protein structure prediction remain unreliable on certain targets. External oracles can flag and correct these failures, but biological oracles are expensive, making oracle budget a critical constraint. Existing guidance methods, such as FK-steering, DPO, and Best K-of-N sampling, differ in how they spend this budget, yet no systematic comparison exists to guide method selection. To bridge this gap, we benchmark these methods alongside the recently proposed Optimisation Over Outputs (O3), which applies off-the-shelf optimisers within a generative model's latent subspace. We extend the usage of O3 to protein structure prediction models. Overall, our work provides the first practical reference for oracle budget-aware guidance. Our evaluation on two protein targets, calmodulin (1CLL) and E. coli aspartate transcarbamoylase (9EEH), reveals that no single method consistently dominates across all budgets and oracles. Specifically, O3 proves most effective at low oracle budgets, while FK-steering and DPO demonstrate improved performance as the budget increases. We distil these findings into actionable recommendations for practitioners operating under real-world oracle-budget constraints.
Read the full abstract
Cryptography and Security
Machine Learning-Based Cyber Defense for Cloud Infrastructure: An Adaptive Deep Q-Network Architecture for Intelligent Intrusion Detection and Automated Threat Mitigation
With the increasing complexity of cyber assaults in cloud environments, adaptable security solutions are needed that can support real-time detection and autonomous response. In this paper, we propose a reinforcement learning-based dynamic cyber defense framework. We deploy a Deep Q-Network (DQN) to train effective defensive strategies to counteract the evolving cyberattacks. We leverage the CICIDS2017 dataset for model creation and the UNSW-NB15 dataset for external validation, involving preprocessing of data, feature engineering, and adaptive policy learning. We compare the proposed DQN with decision tree, support vector machine, random forest, XGBoost, and multilayer perceptron models. The proposed DQN achieves an accuracy of 99.72%, a precision of 99.68%, a recall of 99.65%, an F1-score of 99.66%, and an ROC-AUC of 0.999, while the false positive rate is 0.31%, the false negative rate is 0.35%, and the detection latency is 15 ms. The framework achieved 99.54% attack mitigation rate, demonstrating strong adaptive and real-time defensive capabilities. These results demonstrate the potential of reinforcement learning as a powerful and scalable approach for autonomous cybersecurity in modern cloud environments.
Read the full abstract
Computer Vision
HSTGFormer: Hyper Spatial-Temporal Graph Transformer for 3D Human Pose Estimation
Transformer-based methods have achieved strong performance in monocular 3D human pose estimation, but most existing approaches organise spatial and temporal reasoning as separate stages, which may weaken unified spatial-temporal interdependencies inherent in human motion and compress frame-level structural information before temporal modelling. In this paper, we propose HSTGFormer, a graph-enhanced Transformer framework that reformulates spatial-temporal reasoning as localised coupled graph aggregation over joint-time nodes. Specifically, HSTGFormer introduces a Hyper Spatial-Temporal Graph (HSTG), which decomposes global spatial-temporal reasoning into local spatial-temporal receptive fields around individual joint-time nodes by extending per-frame skeleton graphs into temporal neighbourhoods, thereby enabling structure-aware coupled reasoning while preserving local structural motion information. It further incorporates an Adaptive Dual-Scale Temporal Graph (ADSTG) to capture joint-specific temporal dependencies over complementary short- and long-range windows. A lightweight node-wise fusion module further adaptively integrates the two graph representations for each joint-time node. Experiments on Human3.6M and MPI-INF-3DHP show that HSTGFormer achieves strong accuracy with high computational efficiency.
Read the full abstract
Computer Vision
GenFAR: A generalized representation of brain structure, derived from 49,246 multi-cohort MRIs via deep learning
Deep learning models for neuroimaging have largely been developed for individual tasks, limiting knowledge transfer across applications. Here we introduce GenFAR, a modular deep learning framework that learns general, clinically informed features from brain MRIs. We trained this modular architecture on 49,246 individuals across 11 cohorts, using 17 diverse classification and regression tasks spanning cognition, clinical, diagnosis, demographics, and biomarkers. This yields aggregated, focused feature sets that capture rich, clinically- and biologically-relevant brain representations. We developed a sequential learning approach where tasks progressively build on previously learned representations. Through an analysis of 5,000 task sequences, we identified an optimal sequence length of six tasks and introduced a Donor Score metric to quantify each task's contribution to downstream performance. This analysis revealed five consistently strong donor tasks (Age, AD/MCI, MMSE, Hypertension, Hyperlipidemia) that formed the base of our sequential model. We demonstrated the utility of our learned representation, in various tasks beyond those included in the training set, to serve as the foundation for specialized secondary predictors. We further showed that using the learned feature representation can substantially increase the sample efficiency of secondary deep learning training tasks and models, as well as improve their accuracy.
Read the full abstract
Information Retrieval
Making Collaborative Signals Count: Graph-Aware Large Language Models for Sequential Recommendation
Large language models (LLMs) have been widely adopted as backbones for recommender systems. However, their language-centric pretraining makes it difficult to capture collaborative signals implicit in user-item interactions, which are crucial for personalized recommendation. Existing methods either inject collaborative representations produced by external recommenders or model only intra-sequence dependencies, limiting their ability to exploit global collaborative patterns. To address this limitation, we propose GALLM, a graph-aware LLM framework for sequential recommendation. GALLM constructs a collaborative graph over text tokens and item tokens, and models three types of relations: Text–Text relations for preserving semantic dependencies, Item–Text relations for aligning item tokens with their textual descriptions, and Item–Item relations derived from global item co-occurrence patterns. These relations are transformed into lightweight learnable attention biases and incorporated into the LLM attention mechanism, enabling collaborative-aware token interactions without introducing an additional graph encoder. Experiments on four real-world benchmarks show that GALLM achieves the best performance among the compared baselines, improving over the strongest baseline by 9.76% on average in HR@5. The code is available at https://github.com/Fenglin-miku/GALLM.
Read the full abstract
Computer Vision
Map-Det3D: Metric Feed-Forward 3D Reconstruction Prior for Multi-view 3D Object Detection from Streaming Inputs
Metric 3D object detection is a core capability for embodied agents, yet most reliable systems lean on depth sensors, trading away cost, power, and integration simplicity. This motivates monocular 3D detection, which avoids additional constraints, yet it faces a major obstacle: from a single image, depth, and especially absolute scale, are underconstrained. As a result, the prevailing pattern of detecting in 2D and then predicting 3D attributes is often brittle, since modest range errors can dominate 3D localization, and the learned scale prior can fail when cameras, motion, or environments undergo domain shifts. To address this, we propose Map-Det3D, an online multi-view 3D object detection model that brings detection directly into a 3D space reconstructed from RGB. We map a short temporal window into multiple views and repurpose a feed-forward metric 3D reconstruction model as our geometric backbone while tuning its object-aware capabilities. Building on this representation, Map-Det3D directly predicts boxes in metric 3D space, without the widely used 2D-to-3D lifting. Experiments across different benchmarks show that this design supports strong online performance and robust transfer without adaptation, suggesting that training reconstruction priors for detection is a practical route to stable metric 3D detection from monocular video. Code and models are available at https://royyang0714.github.io/Map-Det3D.
Read the full abstract
Data Structures and Algorithms
Harmonic Ranking for Edge-Weighted Oblivious Matching
We study edge-weighted oblivious bipartite matching. The weight of every potential edge is known, but its existence is revealed only when the edge is probed, and a successful probe between two free vertices must be accepted immediately. We give an explicit randomized algorithm with certified competitive ratio $0.698$, improving the previous best guarantee of $0.659$ (Huang, Sun, Wu, and Zhao, FOCS 2025). The result is computer-assisted and verified by a reproducible exact-integer computation. The same algorithm has a $0.698$-competitive online implementation for the vertex-weighted random-arrival model, improving the previous $0.696$ unweighted guarantee of Mahdian and Yan (STOC 2011) and the $0.686$ vertex-weighted guarantee of Peng and Tang (EC 2025). Our algorithm, Harmonic Ranking, is a role-symmetric generalization of Ranking. It assigns an independent random rank $x_z$ to each vertex and probes a potential edge $uv$ in decreasing order of \[ w_{uv}\frac{h(x_u)h(x_v)}{h(x_u)+h(x_v)}. \] This harmonic priority arises from a budget-balanced gain split and a mutual-proposal interpretation. The analysis lifts two cutoff curves into indicators, reducing the exponential-size factor-revealing problem to a polynomial-size directed minimum-cut instance. A maximum-flow computation with rounded-down integer capacities gives a rigorous certificate. Independently, we observe that the finite-grid unweighted relaxation of our factor-revealing program coincides exactly with a Mahdian–Yan program.
Read the full abstract
Computer Vision
TGRHuman: Text-Guided Realistic 3D Human Generation via Diffusion Renderer
Realistic 3D human generation plays a crucial role in many graphics applications. However, current methods still struggle to generate high-quality human geometry and texture while maintaining 3D consistency and inference efficiency. In this work, we address these limitations by introducing TGRHuman, a novel approach for generating realistic 3D humans from text. Our method decouples geometry and texture generation to alleviate the issues commonly encountered in NeRF-based methods. Instead of relying on slow, implicit score-distillation-based optimization, we directly use explicit multi-view observation generation and optimization for efficient 3D synthesis. For geometry generation, we propose a high-resolution generative module for multi-view normals together with a geometry-carving strategy that preserves view consistency and supports loose clothing. For texture generation, we produce spatially consistent RGB observations from densely sampled surrounding views using a carefully designed texture-prior acquisition strategy and a diffusion renderer, enabling detailed human texture synthesis. Experiments show that our method can generate high-quality and consistent 3D human geometry and texture efficiently. TGRHuman outperforms existing text-to-3D human methods in both geometry and texture quality.
Read the full abstract
Multiagent Systems
Rethinking Agent Security as a Networking Problem
AI agents are rapidly becoming more capable and widely deployed, promising substantial gains in productivity and enabling new classes of applications. However, their growing autonomy also introduces significant privacy and security risks. Existing defenses are predominantly agent-centric, relying on the agent itself to detect threats and enforce privacy and security policies. This approach is fundamentally limited because it entrusts policy enforcement to AI agents whose LLM-driven behavior is inherently nondeterministic and vulnerable to manipulation through attacks such as prompt injection. As a result, current defenses cannot reliably prevent privacy and security threats, highlighting a critical need for a new solution to securing AI agent systems. The networking community has long grappled with similar challenges and offers insightful principles we can borrow to design a more secure AI agent system. These include centralized control with distributed enforcement, capability-based access for mediating requests to sensitive resources, and least privilege through zero-trust enforcement. Historically, these principles have provided strong deterministic guarantees for networked systems. However, these principles alone are insufficient for AI agents because the safety and appropriateness of an agent's actions often depend on semantic context beyond the expressiveness of static rules. Building on these principles, we advocate for a systematic approach to AI agent security that combines deterministic enforcement mechanisms, which provide strong security guarantees, with semantic, context-aware policies that enable nuanced decision-making. We then present a reference architecture and identify key research questions and future directions to guide the design of secure and privacy-preserving AI agent systems.
Read the full abstract
Data Structures and Algorithms
Strongly Polynomial Parallel Maximum Flow Revisited
We study the maximum flow problem in directed networks with real capacities in the parallel setting. For a network with $n$ vertices and $m$ arcs, we show that a randomized parallel implementation of a variant of the strongly polynomial max-flow algorithm of Dadush, Orlin, Sidford, and Végh [SODA 2026] runs in $\tilde{O}(mn)$ work and $\tilde{O}(m)$ depth. This improves upon the previously described tradeoffs between work and depth for strongly polynomial parallel maximum flow algorithms: earlier $\tilde{O}(n^3)$-work algorithms have $\tilde{O}(n^2)$ depth [Shiloach and Vishkin, J. Algorithms 1982; Goldberg and Tarjan, J. ACM 1988], while the known $\tilde{O}(m)$-depth approach uses $\tilde{O}(mn^3)$ work [Orlin, Oper. Res. 1993].
Read the full abstract
Computational Engineering
Physics-Constrained Co-Optimization and Data-Driven Layer-Resolved Classification of a Hybrid CZT/PIPS Detector for Mixed Radiation Fields
Compact mixed-radiation instruments must preserve a low-mass charged-particle entrance while providing enough high-Z depth for photon sensitivity. We first compare two detector heads within a 40 x 20 x 10 mm^3 design budget. S1 places bare CdZnTe (CZT) and passivated implanted planar silicon (PIPS) branches side by side and estimates three rates. S2 adds 0.50 mm of CZT behind PIPS to estimate X/gamma, low-beta, high-beta, and alpha rates. A hard-constraint search combines photon attenuation and deposition, Hecht charge collection, charged-particle energy loss, solid angle, resolution budgeting, timing, and response-matrix conditioning. Feasible screening points exist inside the initial envelope, but a fresh transport/electronics assessment gives only 25.04-25.07 cps/(uSv/h) under the robust H*(10) convention, and every conservative electronics draw exceeds 2.5% FWHM. We therefore retain the more informative S2 observation structure, relax only the 10 mm package-depth constraint, divide the bare CZT into independently biased layers, and add a low-noise sum channel for spectroscopy plus layer-resolved gradient-boosted classification for mixed-field analysis. The extension campaign contains 2.40 million Geant4 11.4.1 histories over three depths, five transport seeds, and 14 particle/energy cases. The final 40 x 20 x 12.570 mm^3 head uses 7.870 mm of bare CZT in five layers at 180 V per layer. Under the application-scenario U95 electronics profile, its live-time-corrected 662 keV sensitivity is 32.804 cps/(uSv/h), its independent-sum resolution is 2.303% FWHM at the 95th percentile, and the propagated beta/alpha absolute-efficiency lower bounds are 35.679%/39.832%. Additional endpoint transport covers 20 keV-3 MeV photons, 3-7 MeV alpha particles, and 155 keV-3.5 MeV beta spectra. The extended design therefore passes the original detector-performance criteria in the U95 model.
Read the full abstract
Computers and Society
Co-constructing sociotechnical AI governance: participatory system mapping using algorithm registers
Algorithm registers have been championed as a means of providing transparency on the use of algorithms in public services. Yet potential publics differ in their expectations of what should be made transparent and how, as well as in their interest in and ability to parse the information currently published in the registers. Moreover, it remains unclear how these instruments can represent the sociotechnical systems in which these algorithms are embedded, and how system-level transparency can facilitate accountability. In this paper, we ask, what do algorithm registers reveal (and occlude) about the sociotechnical systems governing algorithmic systems, and how can diverse stakeholder perspectives inform a more pluralistic system-theoretic safety analysis? To do this, we probe the municipal algorithm register of a Dutch city through a case study of a decision-support tool for caseworkers' assessment of citizens' welfare benefits eligibility based on legal automation through a business rule engine. Through interviews, surveys, and participatory system mapping workshops (with municipal staff, civil society organisations, and ombudsmen, N=8), we seek to understand to what extent the register allows stakeholders to map the algorithmic system in question. These maps inform a System-Theoretic Process Analysis (STPA) that situates the register within a wider sociotechnical governance structure. Participants' contributions allow us to identify potential safety hazards which would not have been possible to see using the algorithm register alone, including benefits eligibility denial, system performance deterioration, and inability to contest wrongful decisions. By engaging both direct and indirect stakeholders, we reflect on the normative dimensions of algorithm governance efforts and how politics shape the practice of system safety analysis.
Read the full abstract
Information Theory
The Optimal Discounting Parameter of the Power Prior under Predictive Log-Loss
The power prior of Ibrahim and Chen incorporates historical data into a Bayesian analysis by raising the historical likelihood to a power $a_0 \in [0, 1]$. The choice of the exponent has remained an open question. This paper gives a closed-form answer under the predictive log-loss. For a model with $d$ parameters, a historical sample of size $N_0$, and average Kullback–Leibler divergence $\bar{D}_0$ between the historical and current data-generating distributions, the optimal exponent is $a_0^{*} = d/(2 N_0 \bar{D}_0 + d)$. Equivalently, the optimally borrowed effective sample size obeys the harmonic law $1/E^{*} = 1/N_0 + 2\bar{D}_0/d$: compatible data are pooled in full, and any difference caps the borrowed information at $d/(2\bar{D}_0)$ observations. The result is exact for multinomial data and extends to smooth parametric families. The law benchmarks adaptive borrowing, explains the reported degeneracy of the normalized power prior, and shows that neither subsetting the data nor decaying the exponent improves on the correctly discounted constant.
Read the full abstract
Computer Vision
Context Blindness in DPO: Mitigating Object Hallucination in MLLMs via Context-Calibrated Preference Optimization
Multimodal large language models (MLLMs) have made rapid progress, yet they still exhibit object hallucination, generating plausible but incorrect descriptions that are inconsistent with the visual input. Direct Preference Optimization (DPO) mitigates this by training models to prefer non-hallucinated responses over hallucinated ones, and recent efforts further enrich the preference data with relevant context. However, it remains unclear whether DPO actually leverages such context. To investigate this, we propose Contextual Preference Gain (CPG), a simple metric that measures how much a model's preference strengthens when relevant context is provided. We find that higher CPG consistently corresponds to lower hallucination, yet standard DPO and its variants exhibit only limited CPG, indicating that they underutilize contextual information and thus remain prone to hallucination. To address this, we propose Context-Calibrated DPO (C$^2$-DPO), which directly maximizes CPG while preserving the original preference ordering. Across multiple benchmarks, C$^2$-DPO substantially reduces hallucination without compromising general reasoning, relatively reducing the Object HalBench hallucination rate of Qwen2-VL-Instruct-2B by 36%. Code is available at https://github.com/mlvlab/C2-DPO
Read the full abstract
Computer Vision
Understanding Why Foundation Models Work for Diffusion-Generated Image Detection
Vision foundation models have recently emerged as powerful feature extractors for detecting AI-generated images, achieving strong generalization across generators and robustness to common image degradations. However, the reason behind their effectiveness is poorly understood. In this work, we investigate what cues are exploited by foundation-model-based detectors to distinguish real images from diffusion-generated ones. To this end, we design an ad hoc analysis protocol based on DDIM inversion. Given a real image we generate a sequence of synthetic copies by changing the depth of DDIM inversion. Even though most copies are semantically identical to the real reference, the detector score varies significantly across them due to subtle traces introduced by the diffusion synthesis, showing that its decision is not primarily driven by semantic failures. Through a frequency-swapping analysis, we further reveal that the discriminative cues exploited by the detectors are mainly localized in the low-to-mid frequency range, rather than only in the high-frequency range, as is the case for artifacts commonly associated with generative models. Finally, a latent-space analysis shows that regenerated images exhibit reduced variance and effective dimensionality, indicating that diffusion models do not fully reproduce the variability of real data. Overall, our results suggest that foundation-model-based detectors succeed by capturing non-semantic low-to-mid frequency distributional discrepancies between real and diffusion-generated images. These findings provide new insight into the robustness and generalization of such detectors and suggest directions for more interpretable forensic methods.
Read the full abstract
Numerical Analysis
Halving the size of skew-symmetric eigenvalue problems via the polar decomposition
This paper introduces a novel algorithm for computing eigenvalues and eigenvectors of a dense real skew-symmetric matrix $A$. Its main ingredient is the computation of a polar factor of $A$ that is both skew-symmetric and orthogonal. This polar factor is then used to transform the original problem into a Hermitian eigenvalue problem of half the size, which can be solved accurately and efficiently with standard software such as LAPACK. Numerical experiments demonstrate the stability of the method and show that its running time is competitive with existing approaches for skew-symmetric eigenvalue problems. Finally, we show that the same principle can be used to reduce an orthogonal eigenvalue problem to a unitary eigenvalue problem of half the size.
Read the full abstract
Artificial Intelligence
Who Thinks Best Depends on How Long You Let Them: Budget-Dependent Rankings in LLM Evaluation
Standard evaluation of large language models assumes stable model rankings across inference conditions. We challenge this assumption by varying the token generation budget, i.e., the maximum tokens a model may produce, across seven levels (64–4,096), evaluating four models on three reasoning benchmarks (56,476 inferences). We report four findings: (i) 3–19% of items exhibit non-monotone behavior (accuracy decreasing with more budget), even after controlling for truncation, and this phenomenon is model-specific (cross-model overlap: 6–14%). (ii) Model rankings reverse across budgets on all benchmarks ($p {<} 0.01$, McNemar). (iii) Oracle analysis reveals model complementarity up to $+27.8$pp, most pronounced at constrained budgets. (iv) A budget-aware router captures 14.1% of the oracle gap cross-domain; budget features help within-domain ($+1.6$ to $+5.7$pp) but are domain-specific and hurt transfer ($-1.2$pp). These results argue for budget-conditioned evaluation protocols.
Read the full abstract
Computation and Language
Massive Activations in Hybrid Linear Attention Large Language Models: Pre-Attention Spikes and Inter-Spike Plateaus
We present the first systematic study of massive activations (MAs) in layer-interleaved Hybrid linear attention large language models (HLA LLMs), examining their architectural organization, training-time emergence, underlying mechanisms, and functional significance. Across five linear attention architectures, six hybridization configurations, and five input domains, we identify two architecture-aligned morphologies: pre-attention spikes (PAS) immediately before full attention and inter-spike plateaus (ISP) persisting through intervening linear attention layers. Denser full attention increasingly connects PAS through ISP, approaching the persistent MAs of conventional Transformers. This organization also recurs across 12 public checkpoints spanning 1.2B-397B parameters, covering linear attention and state-space hybrids. Controlled pretraining of Gated DeltaNet (GDN) hybrids up to 1.3B reveals early emergence and consolidation of both morphologies, alongside asymmetric gating effects. Specifically, full attention output gates strongly attenuate MA magnitudes without eliminating their organization, whereas removing GDN output gates yields modest amplification. Mechanistically, we develop a shared systematic-outlier account: PAS follows a localized write-sink-cancel process, while ISP is consistent with delayed cancellation. Functionally, our interventions show that deleting only the four largest-magnitude PAS coordinates at each full attention input reduces mean downstream accuracy by 21.9%-63.6% relative to normal inference. Moreover, reference-conditioned spike-to-plateau connection consistently improves mean real-world retrieval accuracy, yielding relative gains of 1.1%-12.6% without retraining. Our code is available at https://github.com/StartLuxLabs/Massive-Activations-HLA.
Read the full abstract
Graphics
MVFM-3DAD: Multi-view Flow Matching for 3D Anomaly Detection via Density Proxy Estimation
In 3D anomaly detection (3DAD), most existing methods rely on Memory bank retrieval or reconstruction. However, memory-based methods are constrained by the coverage of stored normal features, while reconstruction-based methods may learn identity shortcuts that also reconstruct anomalous inputs well. These limitations motivate a density-oriented approach that evaluates whether a test sample follows the learned normal distribution. To this end, we propose MVFM-3DAD, a flow-based framework that reframes 3DAD as density proxy estimation over the normal data distribution. MVFM-3DAD introduces a Bidirectional Geometric Projector (BGP), whose forward process converts irregular point clouds into structured multi-view representations. The Flow-guided Density Proxy Estimator (FDPE) estimates a reference density for each view feature, after which the backward process of BGP maps these multi-view density estimates to their corresponding 3D points. Building on it, anomalous features can be identified by their terminal normality. Unlike conventional flow-based likelihood estimation, our formulation requires neither input reconstruction nor explicit Jacobian evaluation, yielding a simple and efficient anomaly-scoring mechanism. Extensive experiments show that MVFM-3DAD outperforms the strongest competing methods on Real3D-AD and MVTec3D-AD. Code is available at https://github.com/lil-wayne-0319/MV3D-AD
Read the full abstract
Distributed Computing
RoutePack: Replay-Guided Expert Placement and Rollout Trajectory Scheduling for MoE Reinforcement Learning
Reinforcement learning (RL) post-training of Mixture-of-Experts (MoE) models couples two load-balancing problems. Rollout-trajectory composition determines attention and dense feed-forward work across data-parallel (DP) replicas, while token routing determines expert work across expert-parallel ranks. Optimizing either side independently can worsen the other. Routing replay exposes trajectory lengths and layer-wise expert demand before training, creating an opportunity to coordinate both controls. We present RoutePack, a replay-guided planner that coordinates expert placement and rollout-trajectory scheduling. Its key insight is that placement determines a batchwise expert-rank load bound that trajectory scheduling alone cannot lower, while scheduling controls the hotspots within individual execution micro-steps. RoutePack incrementally adjusts expert placement while preserving expert residency, then uses parallel population annealing to reschedule complete trajectories into token-capped microbatches. Its objective jointly balances dense-side work and layer-wise expert peaks under a fixed, capacity-feasible micro-step count and the deployed synchronization groups. The planner preserves logical token routing and keeps placement fixed throughout forward and backward execution to maintain training-state consistency. We evaluate mixed-task RL post-training of two hybrid-attention MoE models (7.9B and 124B parameters) across six NVIDIA H20-3e configurations spanning 8K-32K contexts. Within the Megatron training region, RoutePack achieves 4.44-11.85% higher throughput than a DP-local first-fit-decreasing baseline and 1.33-5.28% higher throughput than expert rerouting alone. Additional evaluations of the 7.9B model on H200 and H20 retain positive gains. We separately characterize planning and admission costs.
Read the full abstract
Computer Vision
Autonomous Telerehabilitation via Skeletal Motion Prediction and Joint-Level Performance Assessment
Autonomous rehabilitation systems must not only recognize human motion but also provide structured feedback to support users without continuous therapist supervision. This paper presents a telerehabilitation pipeline that integrates skeleton-based exercise quality assessment and short-term motion prediction into a two-module system operating on marker-free RGB video. A self-attentive Bidirectional LSTM performs exercise quality classification using MMD-NCA metric learning, while a graph-based motion prediction module computes per-joint position errors between predicted and observed poses, generating spatially localized deviation signals. Each module is evaluated independently on established benchmarks: the classifier achieves 96.45% mean-class accuracy on squat sequences from the PROZIS dataset, and the adopted STARS predictor achieves a mean MPJPE of 75.8 mm at 560 ms on Human3.6M, outperforming graph and recurrent baselines across all prediction horizons. The framework is designed for eventual deployment in assistive robotics and home-based rehabilitation contexts; end-to-end integration and clinical validation are important directions for future work. By combining motion recognition and prediction in a single system, this work contributes a step toward autonomous, feedback-driven telerehabilitation, for more accessible and scalable rehabilitation solutions.
Read the full abstract
Software Engineering
ADEPT: A Unified Framework for Deep Learning Test Adequacy
Over the past decade, many test adequacy metrics have been proposed for deep learning that characterize test dataset adequacy from different perspectives, e.g., neuron activation behavior, latent feature coverage, decision-boundary exploration, etc. However, these metrics are typically released as independent research prototypes with substantially different installation and preprocessing requirements, execution workflows, and configuration mechanisms. These complications make them quite difficult to reproduce, compare, and adopt in research work and practical deployment alike. In this paper, we present the engineering details of ADEPT, a framework that integrates representative adequacy techniques, including neuron-coverage-based metrics, surprise adequacy, input distribution coverage, boundary coverage, and source- and model-level mutation score, under a consistent execution workflow. ADEPT provides a template-based metric interface with well-defined extension points for integrating new adequacy metrics. Furthermore, it provides YAML-based configuration management, preprocessing-cache reuse, and structured result reporting, making it easy to use in any research and development workflows. ADEPT is designed for researchers and practitioners who wish to reproduce and apply adequacy metrics without spending days or weeks implementing missing tooling or configuring disparate research prototypes. A demo video is available at https://aub.ie/ADEPT_video.
Read the full abstract
Computational Geometry
Topology-Preserving Meshing of Implicit Scalar Fields via Monotonicity Constraints
Topological analysis of scalar fields yields structures such as the Morse-Smale complex (MSC) that summarize salient features across multiple scales. Existing MSC extraction algorithms typically assume an explicit representation of the input field, such as a discretely sampled mesh. However, recent advances in visualization have popularized implicit field representations, for which these assumptions no longer hold. In this work, we address the problem of extracting an MSC from an implicitly defined 2D scalar field. We present a method for constructing a triangulated piecewise-linear (PL) mesh that aims to preserve the critical points of an underlying implicit scalar field. Our central insight is that if all edges are monotonic with respect to the underlying field, then the resulting PL approximation is topologically consistent with respect to critical points. Based on this insight, we introduce a refinement procedure that mitigates monotonicity violations. Requiring only pointwise evaluations and modest mesh refinement, the approach produces PL meshes that are correct with regards to critical points in our experiments. Finally, we demonstrate that additional targeted refinement improves the geometric fidelity of MSC separatrices.
Read the full abstract
Hardware Architecture
FQTree: Fine-grained Quantization and Hardware Generation of Boosted Decision Trees
Boosted decision trees (BDTs) are widely used in latency-critical applications, but efficient hardware deployment remains challenging. Existing designs often rely on uniform or manually tuned fixed-point formats, which can introduce unnecessary hardware cost or accuracy loss. This work presents the FQTree algorithm{https://github.com/ecs-bristol/FQTree} for fine-grained quantization-aware training of BDTs, together with the QXGB framework for automatic hardware generation. FQTree introduces a hardware-oriented leaf-value quantization scheme that uses a global quantization step together with a tree-wise shift, enabling compact non-negative integer leaf representations, controlled clipping/pruning, and bias folding to reduce datapath cost. This work further applies this quantization during boosting so that later trees adapt to the errors of the already-quantized ensemble, and then lowers the trained model into low-latency hardware implementations through a compiler-based flow. Results on JSC, MNIST, and NID show that our method reduces LUT usage by 26-57% compared with the state-of-the-art FPGA-based BDT designs while matching or improving accuracy.
Read the full abstract
Computation and Language
A corpus-specific clinical RAG system matches or outperforms newer frontier LLMs on HealthBench
General-purpose large language models (LLMs) have recently been reported to match or exceed specialized clinical AI tools on medical benchmarks, but such comparisons draw on a narrow set of systems and on benchmarks developed largely in high-income settings. We evaluate VITA, a retrieval-augmented generation (RAG) system purpose-built for contextual knowledge retrieval in India and other low- and middle-income (LMIC) settings. VITA retrieves from a curated corpus of disease-specific guidelines, India-specific antimicrobial resistance data, national formulary constraints, and resource-limited care protocols; its architecture and corpus are proprietary, but the benchmark, the physician-written rubrics, and our full response and scoring outputs are public for independent verification. On 4,023 English-language HealthBench questions (80.5% of the benchmark), scored with a GPT-4.1 judge, VITA ranked first with 51.9% of possible rubric points, ahead of GPT-5.4 (46.1%), o4-mini (44.3%), Gemini 3.1 Pro (42.6%), and Claude Sonnet 4.6 (37.3%), and scored highest on 45.4% of questions. To test robustness to newer models and judge lineage, a 500-question subset was re-run against current-generation models (GPT-5.5, Claude Opus 4.8, Gemini 3.5 Pro, Grok 4.3) and graded by a neutral open-weight judge (DeepSeek-V4-Pro) sharing no lineage with any system tested. Here the gap narrowed to parity: VITA and GPT-5.5 were statistically indistinguishable on mean per-question score, while VITA led on points-weighted score and won the most questions. VITA's advantages in accuracy and completeness persisted under the neutral judge; its communication scores were lower. These results indicate that a purpose-built clinical RAG system remains competitive with frontier LLMs on an open benchmark, consistent with corpus specificity as a design variable that improves grounding at some cost to communication polish.
Read the full abstract
Game Theory
Tolls for Dynamic Equilibrium Flows
We consider dynamic network flows and study the following question: Which dynamic edge flows can be implemented as tolled dynamic equilibrium flows? We study this question for the heterogeneous-user model, where the flow particles are partitioned into populations characterized by their own source,destination-pairs and a cost function associating with any walk and departure time some costs. As our two main results, we first provide a duality-based characterization of implementability of dynamic edge flows for the multi-source, multi-destination case. Secondly, we derive both, a combinatorial and duality-based characterization of implementability of dynamic edge flows for the multi-source, single-destination case. Both results are derived under a fairly general network loading model. For the proof, we make several technical contributions: We formulate a novel infinite dimensional optimization problem, where the goal is to minimize the aggregated costs of the particles with respect to the fixed network loading induced by the given edge flow. This requires the recently introduced concept of autonomous network loadings for which we show several new structural insights. In particular, we give an alternative (tighter) characterization of the existence of autonomous network loadings for our setting by deriving a generalization of a result of M.A. Zareckiĭ on the Lusin $N^{-1}$ property of absolutely continuous monotone functions which may also be of independent interest. These insights allow us to prove the stated characterizations under the assumption of strong duality. Finally, for the case of a single-destination, we are able to provide a non-trivial proof that this assumption is always fulfilled for finitely supported edge flows with costs representing weighted travel times.