Papers

The ongoing digital transformation of work, administration, health, mobility, and social interaction is profoundly reshaping everyday life, steadily shifting control from individuals to large platform providers. Although data is often labeled the "gold of the 21st century", its real value is realized through services that access, combine, and exploit it. Today, individuals have little sovereignty: life events (e.g., changing an address, insurance, job, or marital status) require fragmented, repetitive interactions across numerous systems, leaving users overwhelmed rather than empowered. We argue for a fundamental rethinking of this situation and propose a virtual, trustworthy platform, called "MyVirtualME" that acts on behalf of the human individual towards companies, administrations, and other actors. This platform centers services, data, permissions, and data usage around humans, not foreign actors, allowing genuine control and transparency. Our vision goes beyond a pure data focus: supported by trusted services ranging from basic notifications to intelligent assistants, MyVirtualME aims to reduce digital bureaucracy, improve and even automate interactions, and support informed oversight of an individual's health, financial, and administrative situation. We firmly believe that a human-centered MyVirtualME is essential for restoring personal sovereignty in our digital future.

The thermal history of the melt pool in laser powder bed fusion (LPBF) additive manufacturing processes governs the solidification microstructure and the mechanical properties of the resulting 3D-printed parts. Dual-laser systems offer additional degrees of freedom to control the cooling profile by reheating material behind the melt pool, but calibrating process parameters is challenging due to the complex physics of the process. We present an optimization-based iterative learning controller that determines optimal power, velocity, and offset settings by judiciously combining simulations and experiments: the model supplies search directions while feedback obtained from experiments on the real plant corrects for parameter errors, enabling convergence despite model inaccuracies. The approach is validated in simulation using a high-fidelity thermal model as a plant surrogate, with deliberate mismatches in absorption coefficient, latent heat treatment, and powder-bed effective conductivity between plant and model. Results show that the controller drives the plant cost down by over an order of magnitude and reaches a tight band of low-cost solutions across seeds, while model-only feedforward optimization stalls at a substantially higher plant cost despite appearing to converge on the surrogate.

Primary bone tumors are rare but clinically aggressive neoplasms whose diagnosis from radiographs is challenged by heterogeneous morphology, subtle lesion margins, and overlapping bone structures. To address the limitations of existing single-view models, we present a dual-input, multi-task learning framework that, to our knowledge, is the first to apply bidirectional cross-modal attention between a lesion crop and the full radiograph for joint segmentation and subtype classification. Using the multi-institutional Bone Tumor X-ray Radiograph Dataset (BTXRD, n=3,746), we employ a YOLO-based detector to generate regions of interest, which are paired with full images as inputs to a dual-stream DenseNet121 architecture. Features are integrated via a novel cross-modal attention fusion strategy, refined by Hierarchical Multi-scale Feature Fusion, effectively balancing fine-grained lesion detail with global anatomical context. Evaluated on a held-out patient-level test split, the model demonstrates superior performance over single-input baselines, achieving an overall Dice Similarity Coefficient of 0.896 and a macro-averaged classification F1-score of 0.928. Notably, the system exhibits exceptional sensitivity for malignant osteosarcoma (AUC 0.999), validating the potential of dual-stream context modeling to support radiologists in accurate, early decision-making.

Approximate convex decomposition (ACD) converts triangle meshes into small sets of convex parts and is a standard preprocessing step for physics simulation, collision detection, and large-scale robot learning. The majority of modern ACD methods produce high-quality decompositions through an expensive search over candidate cutting planes, with per-mesh runtimes of tens of seconds that force game pipelines into overnight bakes and keep articulated-object datasets on CPU clusters for days. Prior work has accelerated isolated stages, most recently VisACD's GPU-based visibility metric, yet the dominant costs – search, mesh cutting, and convex hull construction – have remained on the CPU because their natural decomposition into many small homogeneous phases trails off in a fading last wave at every kernel boundary, and the variable-sized output of each phase forces a host round trip simply to allocate the next launch's input. We address these obstacles by adopting the warp, rather than the thread or thread block, as the unit of algorithm design, an idea introduced in the graph-processing community for a different pathology and which we adapt here to fuse the many heterogeneous phases of a computational-geometry pipeline into single warp-resident kernels, paired with a device-side heap allocator that lets the buffers between fused phases be sized and allocated on the device. Building on this template, we present CuACD (CUDA ACD), the first fully GPU-resident ACD system, together with a suite of reusable GPU components, released as open-source standalone CUDA modules that drop into any search-based ACD pipeline. On the V-HACD benchmark, PartNet-Mobility, and an Objaverse subset, CuACD achieves more than an order of magnitude of speedup over CoACD at matched or better quality.

In recent years a number of finite element methods have been developed for the surface Stokes equations. Constructing standard conforming surface finite element methods for the velocity-pressure formulation of this problem would require simultaneously enforcing $H^1$ conformity and tangentiality of the velocity field, which is not possible on the generalized polyhedral approximating surfaces on which surface FEM are typically posed. Thus methods for this problem weakly enforce either tangentiality or $H^1$ conformity of the velocity field. In this paper we provide an error analysis for an $H({\rm div})$-conforming method based on classical Brezzi-Douglas-Marini elements. In this method continuity of the velocity field is weakly enforced using an interior penalty-type formulation. Advantages of this formulation include exact enforcement of the incompressibility constraint and its construction based on a standard mixed finite element pair. Error analysis of this method in previous work was only valid for lowest-order elements and did not account for “geometric errors”, i.e., variational crimes due to approximation of the continuous surface on which the PDE is posed by an approximating surface on which the finite element method is posed. Here we provide a full error analysis in a natural discontinuous Galerkin energy norm, with geometric errors accounted for. Optimal estimates for the $L_2$ velocity error are also presented. In the case of lowest-order (polyhedral) surface approximations, our proof of optimal $L_2$ estimates requires a stricter than usual condition on the placement of mesh nodes relative to the surface. Numerical experiments are used to illustrate our results.

We propose a multivariable Newton-based extremum seeking scheme, incorporating known bounds on update rates. The design extends the recent bounded extremum seeking scheme to the Newton-based approach, which makes the convergence rate independent of the unknown Hessian by estimating and inverting the Hessian matrix. As a vital part of the Newton-based algorithm, we design an appropriate demodulation matrix to generate an estimate of the Hessian in an average sense. The local exponential stability of the bounded extremum seeking algorithm is proven for general multivariable static maps using averaging analysis. In comparison with previous practical stability results, it guarantees convergence to a neighborhood of the extremum point with an exact convergence rate exhibiting exponential decay. Subsequently, the main stability result is established for the Newton-based bounded extremum seeking approach under a weak Hessian coupling assumption. Simulation results demonstrate the advantage of the proposed approach over bounded extremum seeking by assigning equal, desired convergence rates to all parameters using the Newton approach.

Contextual biasing improves rare-word recognition in speech large language models (SpeechLLMs), but efficiently exploiting large bias lists remains challenging. We propose PTC-Bias, a two-stage framework based on phoneme-level temporal competition. At the prefill stage, PTC Retrieval performs frame-synchronous phoneme decoding and temporal competition among candidate pronunciations, producing a compact bias-word shortlist and corresponding speech intervals. After SpeechLLM decoding, PTC Correction conducts a second local competition between the retrieved candidates and mismatched transcript spans within these intervals. Selective correction reduces near-homophone and word-segmentation errors while preserving correct transcriptions. Both stages share the same phoneme posteriors and require no additional SpeechLLM forward pass. Experiments on LibriSpeech show consistent gains across two SpeechLLMs and bias lists of up to 2000 words. With Prompt-SLAM-ASR-7B and 2000 bias words, PTC-Bias reduces B-WER by 23.4%/23.9% relative to CTC-Filter on test-clean/test-other, while keeping U-WER nearly unchanged.

Machine learning-based Network Intrusion Detection Systems often report near-perfect performance on IoT benchmarks. However, whether these models learn generalizable attack behavior or exploit spurious dataset shortcuts- such as static testbed IP/MAC addresses and chronological recording artifacts-remains an important question. We evaluate the CyberFlowIoT-GICAP benchmark, containing 3,617,388 flow records across 126 PCAP sessions with 849,395 benign flows. Four learning paradigms are evaluated across four feature configurations using PCAP-disjoint splits; LightGBM is additionally evaluated using conventional random-flow splitting. When only statistical flow behavior is used (Fbehav), LightGBM (92.58% +/- 8.18%), Random Forest (92.59% +/- 8.18%), and Deep MLP (92.55% +/- 8.18%) achieve nearly identical Macro-F1, indicating that performance is constrained by feature representation rather than model complexity. With raw timestamps (Ftstamp), tree-based models reach 99.28% Macro-F1, while the linear model remains at 90.62%, showing that nonlinear models can exploit dataset-specific temporal structure. Attack detectability is highly asymmetric: high-rate and active attacks maintain >99.8% recall from flow behavior alone in nonlinear models, whereas the DNS Beaconing drops from 27.78% to 0.00% recall when contextual features are removed. Conventional random-flow splitting increases attack recall by up to 14.00%, highlighting the effect of placing flows from the same sessions in both training and test sets. We conclude with a 4-point protocol checklist for realistic IoT NIDS evaluation.

Imperative code fixes an execution order the computation does not require, and a parallelizing compiler must prove which parts of that order it can remove. We introduce the Canonical Parallel Form (CPF), a device-neutral program state from which every ordering constraint our analyses prove unnecessary has been removed. CPF is reached by output-preserving normalization, by lifting semantic operations such as tensor contractions, and by deriving parallelism in three levels ordered by decidability: syntactic subscript tests, exact affine dependence tests over integer sets, and SMT queries over non-linear integer arithmetic, which also admit parallelism guarded behind a runtime check. Heuristics can then specialize the canonical form for each architecture. Across 248 loop-level reasoning kernels on an AMD MI300A, CPF reaches 4.4x over Numba on its 24 Zen 4 cores and 25.4x on its CDNA 3 GPU. Against the other auto-parallelizing optimizers, CPF is 2.9x faster on the CPU and 8.7x on the GPU than DaCe's own auto-parallelizer, 1.9x faster than Pluto, and 1.5x faster than PPCG on the affine subset of the kernels. Because the pipeline is deterministic, CPF also serves as an agent's starting source, cutting the token cost per kernel by up to a factor of 2.72x while leaving the achieved speed-up unchanged, since the coding agents reason less about parallelism.

Spatial biologists use visualization to assess computational analyses of tissue data. We examine how they direct and verify analyses when an AI agent performs this work. We synthesized workflows from fourteen contextual inquiries and conducted a formative pilot followed by an observational study with ten spatial biologists using Claude Science on their own data. Participants valued help with plotting, locating cells of interest, and tasks they found laborious or could not otherwise perform. Assessing the agent's work involved obtaining suitable evidence, sometimes through additional work in external tools, and interpreting it using knowledge of the tissue and its markers. Scientists also sought information about ongoing computation to decide how analysis should proceed. We contribute a workflow synthesis, an empirical account of scientists directing and verifying agentic analyses, and four design directions addressing execution control, familiar interactive views, source and execution information, and accessible verification across computing setups and experience.

While Federated Learning (FL) has been widely adopted for protecting user privacy in machine learning, it remains vulnerable to various robustness challenges, including performance-impairment risks, information-stealing threats, and aggregation vulnerabilities. This work offers a holistic synthesis of FL robustness along three tightly coupled angles: (i) a threat-centric view of robustness that categorizes the multifaceted attack surfaces, (ii) a structured taxonomy of robust aggregation strategies distinguishing outcome-centric approaches from security-centric strategies, and (iii) a layered taxonomy of defensive strategies. We rigorously examine current evaluation practices for FL robustness and identify major applications and open research challenges to guide future research.

Our mixed-method study examines how Chinese women livestreamers use masking to construct idealized mediated personas while navigating gendered, commercial, organizational, and platform pressures alongside personal agendas. We built on the concept of masking, analyzed 627 recruitment posts, and conducted livestream observations and interviews with 26 Chinese women streamers. We found that streamers assembled masks across bodies, AI-mediated technologies, spaces, performances, and social relations to become recognizable while protecting personal boundaries. These masks were continually negotiated, and participants sometimes broke, resisted, or refused them when demands became misaligned or unsustainable. We conceptualize masking as a sociotechnical assemblage in which agency lies in preserving, disrupting, and reconfiguring relations rather than controlling a single interface. We further theorize breaking as a consequential part of masking that exposes hidden labor and unequal costs of visibility. We offer theoretical and design directions for more negotiable, contestable, and agency-supporting AI-mediated self- presentation.

We study Bayesian linear bandits with an isotropic Gaussian parameter, independent Gaussian candidate arms, and Gaussian reward noise when the horizon is proportional to the dimension. The normalized posterior uncertainty has an explicit limit that is uniform over all causal policies. Gaussian posterior identities then determine the limiting parameter overlaps without an assumed closure of the adaptive recursion. These results yield exact regret curves for Thompson sampling, posterior-mean greedy selection, and a family of policies that scale the posterior sampling covariance. The normalized realized cumulative regret converges in L1, uniformly on compact proportional-time intervals. A policy-uniform lower bound identifies the limiting optimal Bayes regret and proves that posterior-mean greedy selection attains it. Thompson sampling incurs a strictly larger leading regret; its instantaneous regret ratio relative to greedy selection lies between one and two and approaches two at long proportional horizons. Closed-form cumulative curves also identify a different comparison in the vanishing-noise limit. Finally, the instantaneous regret converges to a nondegenerate Gaussian decision-loss distribution, rather than to its mean. The analysis separates the amount of information acquired by a bandit policy from the quality of the decisions made using that information.

Large Reasoning Models (LRMs) have achieved state-of-the-art performance in reasoning tasks by utilizing Chain-of-Thought (CoT) reasoning. To achieve fast execution speed, speculative reasoning techniques adopt a lightweight draft model for candidate token generation followed by process reward models (PRMs) for verification and a strong target model for refinements. This paper identifies that the existing speculative reasoning paradigm follows a strictly forward-only reasoning trajectory, which lacks robustness and can lead to severe error propagation if early reasoning steps are suboptimal. Furthermore, executing these disparate inference schemes, including sequential drafting and parallel verification on homogeneous GPU platforms, can lead to severe resource underutilization. To address this, we propose HeteroReason, an algorithm-hardware co-designed heterogeneous FPGA-GPU inference paradigm specifically tailored for LRM speculative reasoning. At the algorithmic level, we introduce a backtracking-enhanced workflow that enables the system to recover from low-quality states and explore alternative reasoning trajectories, significantly improving reasoning robustness. At the system level, the draft model is offloaded to the FPGA while deploying the PRM and target models on GPUs. A specialized workflow is optimized to achieve prefill-decode disaggregation, which exploits shadow synchronization to overlap GPU-side refinements with FPGA-side token updates to effectively hide synchronization latency. To mitigate inherent sequential constraints, we propose a step-ahead speculation and refinement scheduling scheme, transitioning the system from a sequential execution scheme to a parallel pipeline. Experimental evaluations show an average 4.2% accuracy improvement, with 1.01x-1.42x latency speedups and 1.25x-1.57x improvements in energy efficiency compared to homogeneous GPU baselines.

This paper investigates temporal neural networks for \mbox{end-effector} position \mbox{estimation} of an aerial continuum manipulator (ACM) operating under aerodynamic effects induced by the unmanned aerial vehicle (UAV). An experimental dataset is collected under stationary (\mbox{rotor-off}) and \mbox{free-hovering} conditions across continuum robot (CR) configurations and UAV altitudes, providing \mbox{end-effector} position measurements with and without aerodynamic residuals. To establish a nominal framework, \mbox{strain-parameterized} kinematic models with progressively richer strain bases are evaluated to balance model complexity and prediction accuracy. The selected nominal model then serves as the baseline for 3D position residual estimation using a \mbox{closed-form} \mbox{continuous-time} (CfC) neural network, with a multilayer perceptron (MLP) and a gated recurrent unit (GRU) used for comparison. On unseen test experiments, the CfC achieves an RMSE of \(22.00\pm1.70~\mathrm{mm}\) over five random seeds, compared with \(36.38\pm3.58~\mathrm{mm}\) for the MLP and \(27.72\pm2.92~\mathrm{mm}\) for the GRU, corresponding to reductions of \(39.52\%\) and \(20.62\%\), respectively. These results demonstrate the effectiveness of \mbox{continuous-time} learning for \mbox{end-effector} position estimation under aerodynamic disturbances relative to static and \mbox{discrete-time} learning methods.

Creative and artistic data visualizations communicate stories and invite engagement, yet how designers develop their expressive forms remains poorly understood. We investigate the design process of data artists through three complementary studies: an analysis of 40 public project accounts, artifact-anchored interviews with seven experienced data artists, and a design task-based study with eight data artists. We find that stories and visual forms develop together through data exploration, reference adaptation, sketching, and prototyping. Inspiration comes from data, existing work, everyday imagery, and personal experience. Sketches help develop mappings, while prototypes with real data can reshape representations and intended stories. Designers consider multiple possibilities but typically develop one direction at a time, partly because producing alternatives is costly. Their choices balance meaning, visual appeal, readability, and feasibility. These findings inform tools connecting stories, references, and real data, including AI assistance that supports testing and revising alternatives while preserving designers' creative judgment.

Self-supervised learning (SSL) countermeasures (CMs) have shown strong performance in recent years. However, they often show degraded performance while facing unseen spoofing attacks and mismatched conditions. This study examines the Voxtral audio-language model (ALM) framework for spoofing detection, as a step toward combining CM capabilities within the ALM framework. We analyze how Voxtral captures spoofing cues through audio-text processing and propose an instruction-guided approach that uses label-sequence likelihoods to evaluate bonafide and spoofed speech. Experiments on the ASVspoof databases show that without task-specific adaptation, the LLM layers emphasize semantic representations, reducing the separability of spoof-discriminative acoustic cues compared to the Whisper-based audio encoder. Consequently, spoofing-related information becomes less separable after language-model processing. We also applied lightweight adaptation using weight-decomposed low-rank adaptation (DoRA) to the Voxtral model and propose the Spooftral model, achieving an equal error rate (EER) of 4.25% on the ASVspoof5 evaluation set.

Autonomous UAV flight through cluttered and partially unknown environments requires reasoning not only about observed obstacles but also about occluded regions that the sensor cannot observe. We present OA-MPPI, an obstacle- and occlusion-aware extension of Model Predictive Path Integral (MPPI) control for quadrotor flight that accounts for potential moving agents emerging from these regions into the vehicle's path. At every planning step, we extract a 3D occlusion boundary from the online occupancy map and use it to model the regions that hidden agents could reach over the prediction horizon. We penalize trajectories that enter these expanding regions within MPPI rollouts generated using nonlinear quadrotor dynamics and accounting for individual rotor thrust limits. We validate the proposed approach in simulation and hardware flight experiments, with the complete pipeline running onboard the vehicle in real time. Results show increased clearance from occlusion boundaries compared to baseline MPPI in both settings, as well as avoidance of an agent emerging from occlusion in simulation.

We study the following class of matroid optimization problems with a linear constraint (P-MOL). Given a matroid M=(E,I), two weight functions $v,w:E\to R_{\ge 0}$, and a threshold $L\in R_{\ge 0}$, find $opt v(S)$ where S is either an independent set or a base of M satisfying a budget-type constraint: $w(S)\le L$ or $w(S)\ge L$, and $opt\in\{min,max\}$. P-MOL provides a unified representation for a broad family of NP-hard optimization problems, including budgeted matroid independent set, constrained minimum-basis, and knapsack-cover variants with a matroid constraint. Also, it naturally extends to multiple matroid constraints. In particular, we consider the matroid intersection cover (MIC) problem, where feasibility is defined by the common independent sets of two matroids and one seeks minimum $v(S)$ subject to $w(S)\ge L$. Our main result is a unified EPTAS for all nontrivial P-MOL variants, obtained by generalizing a technique of Hassin and Levin (SIAM J. Comput., 2004) for solving the constrained minimum spanning tree problem. Specifically, for any fixed $\epsilon>0$, we present an algorithm running in time $|E|^{O(1)} (1/{\epsilon^2})^{O(1/{\epsilon})}$ that outputs a feasible solution S whose value is at most $(1+\epsilon)OPT$ for minimization variants and at least $(1-\epsilon)OPT$ for maximization variants. This resolves the complexity status of all members of P-MOL, as none of these problems admits an FPTAS (Doron-Arad, Kulik and Shachnai, ICALP'24). Finally, we separate the P-MOL family from its extension to matroid intersection. We show that an EPTAS is unlikely to exist for the matroid intersection variant of P-MOL under a covering constraint, whereas an EPTAS is known to exist under a budget constraint. This highlights a qualitative difference between these two types of linear constraints that does not arise in the single-matroid setting.

Hate speech on social media poses serious risks to social harmony, mental well-being, and public safety, making its timely and accurate detection essential for content moderation systems. Most existing studies focus on binary classification, evaluated their frameworks on a single dataset, and provide limited insight into how decisions are made, which limits their real-world applicability. In addition, limited work is done on the explainability of their predictive inference. To address these challenges, this study proposes a multilevel and explainable hate speech detection framework. The proposed model integrates DistilBERT (Distilled Bidirectional Encoder Representations from Transformers) embeddings with a Bi-LSTM (Bidirectional Long Short-Term Memory) model, and an attention mechanism to capture both contextual meaning and sequential dependencies in text. To enhance trust and transparency, LIME (Local Interpretable Model-agnostic Explanations) is employed to explain model predictions by highlighting influential textual features. The framework is evaluated on two benchmark datasets using both binary and multi-class classification to examine robustness and generalization. In addition, an ablation study is presented to highlight the significance of various components of proposed framework. For binary classification, the proposed model achieves F1-scores of 96.78% on the Davidson dataset and 99.53% on the SMHS dataset. In the multi-class setting, it attains F1-scores of 97.00% and 94.99% on the Davidson and SMHS datasets, respectively, outperforming existing baseline approaches. The results demonstrate that multilevel evaluation improves the reliability that the proposed framework effectively balances performance and efficiency. This makes the framework suitable for practical hate speech moderation systems that require accurate, generalizable, and explainable decisions.

We study square-complementary graphs $G$ (satisfying $G^2 \cong \overline{G}$). We show that in such graphs, no two vertices have comparable closed neighborhoods. This implies that nontrivial square-complementary graphs have no simplicial vertices and are not chordal, thus solving two open problems posed in Discrete Mathematics 327 (2014) 62-75. We also show that no nontrivial square-complementary graph is distance-hereditary.

Dynamic Spectrum Access (DSA) through the Spectrum Access Systems (SAS) elevates spectral efficiency, yet existing centralized models face allocation logic opaqueness and a lack of independent verifiability. While blockchain-based SAS architectures offer transparency and verifiability by default, they introduce critical privacy risks and prohibitive on-chain computational overhead. We introduce zkSAS, a practical zero-knowledge proof (ZKP) system designed to address the verifiability and privacy gaps in SAS deployments, with direct applicability to both the existing CBRS SAS model and blockchain-based SAS models. The system features a suite of ZKP circuits, encompassing proofs of allocation constraint validity and proofs of move list validity to verify that channel assignments and move list-based incumbent protection measures, respectively, adhere to regulatory constraints without exposing sensitive user data. Comprehensive evaluation of our prototype in both centralized and blockchain-based settings indicates that while proof generation scales with spectrum user population, verification remains lightweight and constant-time. We envision that zkSAS offers a scalable and practical path to secure, verifiable dynamic spectrum sharing.

We present a safe receding-horizon mixed-integer differentiable predictive control methodology for residential battery energy storage dispatch that combines neural-network speed with recursive feasibility guarantees. Unlike open-loop learning-to-optimize methods, it incorporates real-time state-of-charge feedback and sinusoidal time-of-day conditioning, enabling closed-loop re-planning at every timestep without re-solving a mixed-integer program. A differentiable rainflow cycle-counting layer enables self-supervised training of the mixed-integer policy on exact degradation physics. The controller is a hybrid closed-loop system pairing a neural mode-selection and continuous-action policy with a quadratic-programming safety filter that guarantees recursive feasibility independent of network weights or mode optimality. We establish mode-conditioned Lipschitz continuity and a conditional regret decomposition into training-quality, mode-mismatch, and forecast-error terms. On a 7-day net-metering evaluation, the method attains a 6.9% cost gap versus the closed-loop mixed-integer MPC benchmark with a 25x speedup (0.11 s vs 2.7 s per step), while average regret rises by under 4% across 0-30% forecast noise. The learned and benchmark modes agree at every step, so the bound reduces to its training-quality and forecast-error terms, both empirically validated.

Scientific tasks specify a desired capability, but realizing it often requires building a computational system tailored to the task—acquiring data, designing representations, training models, implementing tools, and deciding how they are used at inference. We present, a framework in which an AI builder turns a scientific brief into an executable AI lab: a task-specific solver that integrates models, knowledge resources, tools, and a controller behind a fixed interface. The builder develops and packages the lab in a metered workspace; a separate host then executes the delivered artifact on held-out inputs, with reference labels kept outside the solver's input interface, and scores its outputs under the task's protocol. This makes the delivered system, rather than the builder's account of its progress, the object of evaluation. We document 10 selected constructions across six scientific task categories—from molecular and genomic prediction to medical imaging, clinical decision support, and biomedical text—whose delivered labs exceeded their configured reference values on all 12 subtests under host-side execution. Four contain predictive models fitted during construction; the others assemble executable analysis environments, knowledge resources, and tool-driven workflows around a fixed platform LLM. Together they show that an AI agent can carry a scientific brief all the way to a working lab that can still be invoked, inspected, and checked after construction ends.

This work presents a federated framework for training Averaged $n$-Dependence Estimators (AnDE) in distributed environments. The proposed method focuses on the discriminative setting, where model weights are learned locally and aggregated globally, supporting any dependency order $n$. This design allows federated training without transmitting semantically meaningful parameters, improving privacy. Additionally, generative AnDE models are federated to provide a comparative baseline, with optional differential privacy applied to the aggregation of probability tables. Experiments on 12 discrete datasets show that discriminative models with $n \geq 1$ consistently outperform federated Naive Bayes (NB, $n=0$), and that privacy-preserving aggregation is effective with limited accuracy loss. These results establish federated AnDE as a viable and privacy-preserving framework, showing that probabilistic models remain applicable in modern federated learning settings.

Large Language Model (LLM) agents struggle to scale safely when exposed to vast enterprise toolsets. Providing an agent with access to every internal tool leads to oversized context windows, degraded tool selection, and severe governance vulnerabilities - as system policies defined purely in prompts remain probabilistic advice rather than hard constraints. Existing mitigations, such as multi-agent domain delegation, decentralize audit logs and fail to guarantee policy compliance across sessions. We introduce skilder, a framework that packages capabilities into roles: bundles of skills, tools, and instructions, together with the limits that bound them. An agent begins with a minimal role catalog, learns the roles a task requires, and receives each role's skills, instructions, and tools through a single MCP server. Because tools reach the agent only inside learned skills, the same server enforces the scope of what was learned deterministically. We evaluate skilder against flat-context tool selection and multi-agent orchestration across 13 tasks using six models (10 runs each). Our results show that, when models completed discovery and issued a governed call, the skilder simulated authorization layer enforced governance boundaries: no unauthorized tool call or parameter violation (e.g., a spending-limit breach) executed. Aggregate task pass rates also reflect whether each model followed the discovery protocol and satisfied response-quality checks; those misses are not authorization failures. Furthermore, by allowing agents to dynamically acquire cross-role capabilities mid-task, skilder preserves problem-solving flexibility while providing hard system-level enforcement.

Artificial Intelligence agents based on large language models provide convenient natural language interfaces to scientific software, but reliability is not automatic. Here we introduce the Epydemix Agent Framework, an additive layer over Epydemix, an open-source Python library for stochastic compartmental epidemic modeling. The framework extends the library with four capabilities to facilitate interaction with an AI agent: discovery of available models and parameters, preventive validation of a declarative scenario specification, execution through tested library code, and inspectability of results. These capabilities let an agent handle the entire modeling process, from the natural-language description of the scenario to quantitative results, figures, and interpretation of findings without writing custom code. Each step reads input files and saves results in a separate output bundle, making the process auditable and reproducible. First, we show the end-to-end workflow with a case study comparing vaccination strategies for a novel respiratory virus. Second, we assessed the framework across 50 agent sessions and five modeling tasks by comparing the agent use of the framework against the direct use of the Python interface. The framework reduced turns, output tokens, and cost on most tasks, unless it trades resources for per-point reproducibility.

A communication problem can have far more possible inputs than its communication cost would suggest. Must its difficulty already be present on a much smaller set of inputs? We prove that every finite total Boolean matrix of deterministic communication complexity $c\ge4$ has a submatrix on $2^k$ of its original rows and $2^k$ of its original columns, with $k=\Theta_\varepsilon(c)$ and complexity at least $(1-\varepsilon)(k+1)$, for every fixed $0<\varepsilon<1$. Since $k+1$ is the maximum possible cost on such a square, the retained problem can be arbitrarily close to maximally hard. This answers affirmatively the lossless condensation question of Hamed Hatami; Göös, Newman, Riazanov, and Sokolov (STOC 2024), who recorded it as Open Problem 2, conjectured a negative answer. Hrubeš previously guaranteed input length $\Omega(\sqrt c)$. The same argument gives an original $2^{c-2}$-by-$2^{c-2}$ square retaining at least $c/3-O(\log c)$ bits of communication complexity. The proof builds on Hrubeš's counting and covering argument. We count submatrices equipped with short communication protocols: a player names a covering submatrix, then the players run its protocol. This avoids the loss from converting rectangle partitions into protocols. A recursion on rectangles makes the argument constructive. For fixed rational $\varepsilon$ and any target depth $d\ge4$, a deterministic algorithm returns either a protocol of depth below $d$, or a square of original inputs at input length $\Theta_\varepsilon(d)$ with the same near-maximal guarantee. Its running time is $2^{O(2^d)}$ times a polynomial in the table size. If the original complexity is at least $d$, the algorithm necessarily returns the square. An extension gives constant-factor condensation for any fixed number of number-in-hand players, with bounds independent of the finite output alphabet.

Faithful user simulation is fundamental to building, evaluating, and improving interactive AI at scale. Yet current simulators often produce plausible individual responses without reproducing the intent evolution and outcomes observed in real interactions. We propose TRACER, a multi-turn user simulator that models evolving user intent and aligns simulated trajectories with real ones. TRACER is trained in two stages: supervised fine-tuning on real user dialogues, followed by multi-turn reinforcement learning. The RL stage combines hierarchical outcome- and trajectory-level rewards with deviation-aware advantage modulation, jointly addressing reward sparsity and credit assignment challenges in long dialogues. On real customer-service sessions organized into reference cohorts, TRACER-7B surpasses the strongest baseline by 11.4 conversion F1 points, while outperforming all baselines on group-level conversion-rate error and semantic trajectory distance and generalizing to out-of-distribution scenarios. In human Turing tests, annotators identified TRACER conversations at near-chance accuracy. Building on this simulator, we further introduce the Dynamic Marketing Benchmark, which jointly evaluates persuasion and response quality via simulated interactions, revealing that higher response quality does not necessarily correspond to higher conversion rates.

Social media platforms are embedded in teenagers' daily lives, supporting friendship and identity while exposing teenagers to unwanted contact and privacy harms. Previous scholarship has documented how attention capture strategies and dark patterns shape social media use, and we extend this work to better understand platform settings that ostensibly provide privacy and safety protection. We report on think-aloud sessions with 11 teenagers aged 14 to 17 who completed six privacy and safety tasks on Instagram, TikTok, Snapchat, and YouTube. We show how participants worked out what a setting meant through their routines, boundaries, and prior experiences, how they accommodated protections softer and less predictable than expected, and how they treated the platform as the authority on what protection should look like. We argue that feature-by-feature evaluation cannot establish whether teenagers are protected, and that platforms should carry the obligation to show that a protective action took effect and is durable.