Papers

Immigrant-led nonprofit groups, particularly those operating in politically sensitive contexts, face exclusion from formal registries and digital platforms. This paper reports a three-phase mixed-methods study with Iranian immigrant nonprofit practitioners: 27 semi-structured interviews, a co-design session, and 7 evaluation and feedback interviews on a prototyped AI assistant, AMINA. Our findings highlight how legitimacy barriers, capacity gaps, and politically charged misinformation constrain nonprofit operations. We translate these insights into design goals for an inclusive nonprofit AI assistant: support for everyday group operations, recognition of informal nonprofit efforts, proactive countering of misinformation, and multilingual, accessible interaction. User evaluations show AMINAs potential to reduce reporting burdens and foster transparency through proactive reminders, and catalyze collaboration across dispersed networks. We contribute to CSCW and HCI by characterizing the cooperative work of transnational immigrant nonprofits, extending scholarship on informality and misinformation, and demonstrating how AI can act as a collaborative partner that strengthens, rather than displaces, the human connections at the core of nonprofit ecosystems, while also posing major risks.

Multi-agent AI platforms move quickly from staging to production, but the way agents establish trust remains rudimentary: an agent either transmits raw data to a peer or accepts that peer's natural-language self-report that a value complies with policy. The first over-shares; the second is unverifiable and is exactly the channel prompt injection attacks. Prevailing responses emphasise identity, visibility, and post-hoc detection, and recent proposals for cryptographically enforced agent policy have been evaluated in simulation rather than execution. We take provable data minimisation between agents from proposal to running system. In our Zero-Knowledge Proof Gateway, agents exchange proofs of governance-defined predicates over private data rather than the data itself, so exposure is prevented by design rather than detected afterwards; because no interoperability protocol can carry such a proof, we propose a slot and implement it on both MCP and Agent2Agent from one endpoint. A 32-bit threshold predicate proves in 6.2 ms and verifies in 1.0 ms with a 608-byte Bulletproofs proof on one commodity vCPU; eleven adversarial experiments and nineteen protocol checks pass; and the system is deployed to Kubernetes with empirically verified network isolation. Our case study proves a retail client order is within its limit without revealing the amount, instantiating the GDPR data-minimisation principle as an enforced technical measure of the kind EU law now names explicitly. We then address the limitation no comparable work resolves: a predicate proof binds a statement to a committed value, never to the system of record. We give a construction fusing an enclave attestation with the proof in both directions, so verifying one artifact certifies jointly that the predicate holds and that the value was read by a specific measured binary, and test it against a mock authority.

Anti-Latin squares were introduced in connection with non-linear secure network coding, and the extremal problem for large mutually orthogonal families is motivated by that setting. We study the maximum size $N_A(d)$ of a family of mutually orthogonal anti-Latin squares of order $d$. We prove that $N_L(d)+1\le N_A(d)\le N_L(d)+2$ for every $d\ge 3$, where $N_L(d)$ denotes the classical maximum size of a family of mutually orthogonal Latin squares of order $d$, and we show that in fact $N_A(3)=N_L(3)+1$ whereas $N_A(d)=N_L(d)+2$ for every $d\ge 4$. The upper bound is obtained by passing through balanced matrices, while the lower bound is given by a deterministic permutation argument. For all $d\ge 8$, and also for the exceptional order $d=6$, the upper bound is shown to be attainable by a general probabilistic construction. On the structural side, we show that a saturated family of size $d+1$ induces an affine plane of order $d$, and that the saturated case is characterized by the existence of an anti-coordinate grid decomposition; after transporting this condition to the fixed cell set $[d]^2$, it becomes a direction-completeness condition on the corresponding row-blocks and column-blocks. The remaining small orders are treated separately: $d=3$ is handled by direct analysis and classification of orthogonal triples, $d=4$ by an explicit saturated construction and an analysis of its finite-geometric structure, and $d=5$ and $d=7$ by explicit saturated examples arising from the random-grid framework. Thus $N_A(d)$ is determined in terms of $N_L(d)$ for every $d\ge3$, and its numerical value is obtained explicitly for every $3\le d\le9$.

Understanding which training samples influence a generated image is an important problem in generative modeling. In flow matching, training samples influence the generated image through the velocity field along the generation trajectory. Removing samples to examine their counterfactual influence changes the velocity field, and the resulting effect on the final image depends on how the change propagates through the trajectory. Consequently, local changes in the velocity field do not necessarily predict the final counterfactual effect. This work investigates attribution in flow-matching models through a hybrid analytical–learned approach, and uses it to derive trajectory-based attribution scores at the cluster level. We evaluate these attribution scores using independently retrained leave-one-cluster-out (LOCO) models, and compare with several attribution baselines using two different flow-matching latent spaces. Our experiments show that semantic similarity constitutes a strong baseline, while the closed-form trajectory-based attribution is competitive in some metrics without requiring counterfactual retraining or model gradients. Our results show that attribution in flow matching depends not only on semantic similarity to training samples, but also on the latent representation, trajectory dynamics, and how influence is propagated to the final output.

Single-GPU deployment of 70B-parameter language models on an NVIDIA GPU is constrained by device memory, long-context throughput, and engineering integration cost. We cast single-GPU inference as a budget-aware design problem over these three axes and study how pruning, quantization, and KV-cache compression interact under realistic execution. Controlled ablations show that layer-wise pruning makes weight quantization more robust. KV-cache sparsification complements INT8 KV quantization by reducing memory without hurting decoding speed, while static vector quantizers often conflict with dynamic caching. Guided by these coupling results and explicit budget tracking, we assembled a practical pipeline and compressed a 70B model to about 33 GB, sustained about 57 tokens/s on 10k token prompts on a single A40, and kept absolute accuracy within 5% on common and reasoning benchmarks. We contribute design rules and a reproducible evaluation protocol that jointly report quality, memory, and end-to-end speed, and we provide a foundation for automated pipeline search under realistic single-GPU constraints.

Token prediction is a central pre-training objective for modern language models. Despite its empirical success, why token prediction learns broadly useful representations remains incompletely understood. We develop a statistical framework connecting token prediction with representation geometry, encoder approximation, and downstream performance. Under a softmax prediction head, we show that accurate token prediction organizes token embeddings according to similarities between the distributions of contexts in which different token types appear, as measured by Hellinger distance, with explicit errors governed by prediction accuracy and token frequency. Meanwhile, the contextual representation provides a low-dimensional coordinate for the conditional distribution of the target token relative to these embeddings. We further introduce a self-consistency principle showing that repeated applications of a shared representation block can progressively refine the contextual representation without introducing additional block parameters. Among representations with the same prediction accuracy, this recurrent construction favors those that can be stably reconstructed from their contexts. Finally, we establish downstream guarantees for token generation, token community recovery, and classification by a linear probe, showing how prediction accuracy and recovered geometry translate into performance beyond the pre-training objective. Together, these results explain how the simple objective of predicting tokens can recover semantic geometry and produce broadly useful representations. A controlled simulation illustrates the theoretical mechanisms.

Sparse representations are often expected to make models smaller and also reduce inference cost. For Fourier Neural Operators (FNOs), these objectives are not equivalent or do not always align: removing parts of the learned operator can leave the underlying transforms and dense computations unchanged, while changing the grid on which the model is evaluated can introduce overhead of its own. We therefore distinguish sparsity in the representation, in the stored parameters, in the theoretical operation count, and in measured runtime, and present an empirical study of several routes toward sparse FNOs that tests each transition between them separately. Coarsening the execution grid reduces the theoretical cost without reducing measured latency, and adding a correction term recovers accuracy at the cost of making the model slower. Even an 83% parameter reduction remains slower than the dense baseline under ordinary execution. These results motivate a stricter definition of useful sparsity: the deployed operator must preserve solution accuracy and map its reduced support to a genuinely cheaper execution path.

Films communicate through deliberate creative choices, including lighting, color, composition, editing, dialogue, music, and sound. Humans naturally interpret these signals as directorial intent, yet current multimodal large language models (MLLMs) are evaluated almost exclusively on understanding what happens rather than why it is presented that way. We introduce TAKE 85, the first benchmark for directorial-intent understanding, comprising 398 short films (85 hours) with expert-verified question-answer pairs spanning global and fine-grained visual and audio intent. Through controlled modality ablations, TAKE 85 enables systematic evaluation of multimodal reasoning. Experiments on state-of-the-art MLLMs reveal a substantial gap between perceptual recognition and intentional understanding: while models accurately describe events and narratives, they consistently fail to infer the communicative role of filmmaking decisions. Our results establish directorial intent as a previously overlooked dimension of multimodal understanding: even the strongest model reaches only 58 out of 100, and our ablations show that no input modality is sufficient on its own. All code, Q&As, and models are publicly available from https://github.com/KaiShinozakiConefrey/Take-85

How do LLM agents come to both understand environments they act in and master tasks set within them? Through controlled experiments combining world-model training (next-state prediction) and policy training (reward maximization), we investigate this question. We dissect the resulting models through their additive parameter updates. Geometrically, we find effective world-model updates are low-rank and share an input-feature subspace with policy updates while writing to nearly orthogonal output directions, whether trained separately or sequentially. However, we find that, in projection interventions, the sequential update induces more robustness than separate policy RL when removing the world model's leading input directions, suggesting that it has learned alternative input pathways. Behaviorally, we find the sequentially trained agent explores a wider range of states and actions. Based on this, we ask: does policy training preserve world knowledge as well as it could? We probe this with training-free merging built on the geometrically motivated input basis plus an online world-model loss during policy RL, and show both improve over the untreated baseline. Our findings suggest world knowledge and task-directed ability can be learned in geometrically complementary forms, and that future post-training pipelines should consider how best to engineer the interface between them.

Autonomous trucks (ATs) may introduce distinct pedestrian-safety risks because of their large physical dimensions, constrained braking capability, limited driver-based communication cues, and potential to occlude surrounding traffic. This study employed a controlled virtual reality experiment with 54 participants to investigate pedestrian-AT interaction risk in an unsignalized multi-lane crossing scenario and to evaluate occlusion-targeted risk mitigation strategies. The experiment examined the effects of near-side vehicle type, weather condition, and far-side vehicle yielding strategy on pedestrian behavior, perceived risk, and objective safety. Based on a representative high-risk scenario, three targeted interventions were designed and tested: an environment-aware external human-machine interface (eHMI), a projected eHMI, and an auditory warning. The results showed that ATs increased perceived risk and encouraged more cautious crossing behavior, suggesting a risk-compensation effect. However, this compensation was weakened under rainy conditions, where braking-related safety margins were reduced. AT-induced occlusion further increased far-side interaction risk by limiting pedestrians' recognition of hidden vehicles. Among the three interventions, the projected eHMI showed the best overall performance, improving objective safety margins, enhancing risk awareness, and supporting behavioral adjustment. These findings highlight the need for AT-specific interface and warning strategies that address both intention communication and risk localization.

Large language models (LLMs) demonstrate strong Helpfulness, Harmlessness, and Honesty (3H) alignment in English-centric settings, but these gains transfer poorly to low-resource languages due to cultural mismatches. Existing multilingual 3H benchmarks rely predominantly on automated translation or LLM based synthesis, propagating source-language biases while sacrificing local relevance. To address this gap, we introduce Pak3H1, the first human-validated, culturally contextualized Urdu benchmark suite for 3H alignment, comprising PakAlpaca (helpfulness), PakBeaverTails (harmlessness), and PakTruthfulQA (honesty). Our multi-stage pipeline integrates manual cultural adaptation and dictionary-guided post editing to prioritize native speaker judgment, ensuring both semantic fidelity and contextual authenticity. Zero-shot evaluations across multiple open and proprietary LLM architectures reveal systematic cross-lingual alignment gaps: helpfulness win rates decline under localized contexts, harmlessness guardrails break down against regional safety risks, and composite honesty metrics degrade substantially due to localized factual constraints. These findings expose structural limitations in current alignment approaches, underscoring the necessity of human-guided localization for equitable multilingual evaluation.

We develop an immersed interface finite difference method for a one-dimensional nonlinear parabolic interface problem with jump condition \[ [u]_\alpha=\lambda u^+u^-. \] The method combines a Crank–Nicolson immersed interface discretization with an \(s\)-parameter reduction of the nonlinear interface condition, thereby reducing the nonlinear coupling to a scalar quadratic equation. We also discuss two different viewpoints for combining Newton iteration with immersed interface discretization, namely the IIM–Newton and Newton–IIM formulations. Numerical experiments are presented to illustrate the behavior and accuracy of the method.

Adiabatic logic has traditionally been recognized as a low-power solution but constrained to low clock speeds to preserve adiabatic behavior. For advanced FinFET nodes, however, clock frequencies have plateaued due to power/thermal concerns (dark silicon) even as the intrinsic device speeds have continued to scale. This convergence opens an opportunity for adiabatic logic to maintain adiabatic behavior even at GHz clocks. We demonstrate an adiabatic logic (AL) design methodology through a MAC systolic array implemented in commercial 16 nm FinFET technology with a resonant 4-phase power clock (PCK) generator, including digital-to-AL and AL-to-digital interfaces. Simulations show that the AL MAC systolic array at 1 GHz achieves power reductions of up to 42% and 36% at the core and system levels, respectively, compared to digital counterparts. Scaling to more advanced nodes should provide even better power/performance metrics.

JSON Bag-of-Tokens (JSON-Bag) is a recently proposed method to generically represent game trajectories by tokenizing their JSON descriptions. We introduce JSON-Bag VF, a game-agnostic approach to training value functions for game-playing agents using JSON-Bag prototypes. We show that this approach can be enhanced with Random Forest-based feature selection and a method to select game-stage-specific features. We evaluate JSON-Bag VF with One-step-look-ahead (JSON-Bag OSLA) on six tabletop games over different combinations of prototype-tokenization and feature selections. JSON-Bag OSLA outperforms baseline OSLA agents in most games. Our analysis also shows that feature selection significantly improves JSON-Bag VF and that feature selection is the most important factor in JSON-Bag VF performance, over prototype-tokenization.

Massage therapy is helpful for the rehabilitation of various diseases, such as headaches caused by migraines and stress. Existing robotic systems have focused on massage therapy on the torso and limbs, but performing massage motions through suitable actuation on a person's head has been a challenge. In this paper, we present MiBOT, a head-worn massage robot that actuates two soft tactors to produce touch motions mimicking human massage. A key design principle behind MiBOT is its silent actuation, which we achieve through pneumatic artificial muscles in conjunction with a controller loop to respond to contact pressure. We evaluated the effectiveness of MiBOT in a controlled study and assessed subjects' blood pressure and heart rate levels while applying MiBOT. We found that our mechanical system generated positive and conclusive quantitative outcomes that are similar to the human-administered massage, decreasing participants' mean systolic and diastolic blood pressure by 2.8 mmHg and 1.7 mmHg, respectively, as well as calming their heart rate by 8-10% on average.

When does 3D Gaussian Splatting (3DGS) recover the true scene surface rather than just overfitting view-dependent appearance? We answer this by developing a mathematical framework based on a first-hit rendering abstraction that cleanly isolates geometry from appearance. We prove that geometric misalignment forcefully converts spatial textures into high-frequency angular signals via parallax. This establishes a strict identifiability window: if angular capacity is bounded, surface-consistent solutions are mathematically preferred; if unrestricted, the same images can be perfectly explained by an incorrect, opaque billboard geometry. Experiments on synthetic stress tests confirm this prediction, showing billboard failures emerge precisely at high angular capacities. Conversely, in the real-world datasets we evaluate under standard capture protocols, reconstructions remain surface-consistent even at high SH degrees, which is consistent with the prediction that rich spatial texture can push billboard solutions outside the tested angular-capacity range.

When a generative search interface answers a commercial question, which market's products it names is decided before the model reasons about the products. We report a controlled probe of 234 runs against the logged-out ChatGPT web interface and the OpenAI API, collected on 29 and 30 August 2026 across four exit countries and six query languages, with six identical runs per cell. Three results. First, the top recommendation is unstable: it changed across six identical runs on four of six prompts, and that rate was identical in the browser interface and in the API with web search both enabled and disabled, so instability is a property of the system and not of the surface. Second, query language, and not location, decides whether local suppliers appear at all. Where the query language matched the country, a global brand won 1 of 24 runs; asked in English on the same connections, local brands took 0 of 6 runs in Estonia and Turkiye. Third, language and location are separable and act on different things: holding the query language fixed and moving only the exit IP moves the market whose brands are named while the answer stays in the query language. We show this on two unrelated pairs, Turkish asked from Berlin and Russian asked from Tallinn, and in both the answer names the resident country's suppliers. A minority language occupies a middle tier: Russian asked from Estonia names an Estonian supplier in 4 of 6 runs and a global one in all six, where Estonian names a local supplier in every run and English names none. A negative control in a second category, coded with the same instrument, shows no language effect at all, and disconfirms our own expectation: that category does have domestic suppliers and none was named in any language, which points the explanation at whether a category is nationally regulated rather than at whether it is nationally supplied.

Process reward models (PRMs) provide dense step-level guidance for search-based reasoning, enabling inference-time compute to be allocated toward promising partial solutions. However, recent evidence suggests that PRM-guided search can over-optimize imperfect process rewards, pruning viable trajectories while expanding spurious ones. In this work, we theoretically show that directly leveraging PRM score is vulnerable to verifier noise through an extreme-value effect: non-viable prefixes become more likely to receive spuriously high scores as reasoning depth increase. Therefore, we formulate the PRM-guided search as a robust optimization problem over plausible reward perturbations, termed maximin PRM-guided search, leading to a training-free robust process supervision method that preserves promising alternatives when step-level scores are noisy. Maximin PRM-guided search mitigates this failure mode by reducing sensitivity to over-optimized PRM outliers. Without fine-tuning or online adaptation, maximin search consistently improves the PRM-guided search by 17-35% on average, outperforming outcome- and step-level baselines in 14 out of 16 settings. Our source code is available at https://github.com/tjoo512/maximin-search.

Building a blockchain digital twin largely requires translating domain knowledge and specific system descriptions into a simulator architecture, calibrating its parameters against behavioral evidence, and validating the constructed twin. These steps are commonly performed through application-specific modeling efforts that can be difficult to reuse across systems and downstream decision problems. We consider automating this process through Spec2Twin-Chain, a framework that formulates blockchain digital-twin construction as a bi-level optimization problem. At the upper level, a large language model proposes and revises structurally admissible architectures using system specifications, behavioral evidence, and feedback from evaluated designs. At the lower level, a simulation-based optimizer calibrates the architecture-conditioned parameters under explicit objectives and guardrail constraints. The two levels iterate. The evaluated candidates at lower levels are retained in a global archive and used to guide subsequent proposals at upper levels. We conduct controlled experiments involving twin calibration, feedback-driven recovery, stress analysis, downstream policy optimization, and policy updating. The results demonstrate that the framework can construct behaviorally accurate twins, improve initial designs through iterative feedback, and reuse calibrated twins to support downstream decisions.

Recent AI systems promise autonomous scientific discovery, claiming to discover algorithms and produce research papers, yet understanding whether they exhibit creativity, the capacity to produce solutions that are both novel and useful, remains an open question. We present a framework for evaluating multi-turn LLM research agents' creativity using ML engineering tasks as a testbed, through three dimensions: P-Creativity (psychological novelty: novel relative to the agent's own prior solutions within a run), H-Creativity (historical novelty: novel relative to the corpus of human solutions), and Usefulness (task performance). Evaluating two agent frameworks, AIDE and AIRA-Dojo, on 10 Kaggle-style machine learning tasks from MLE-Bench, we develop an LLM-as-a-Judge pipeline and verify its strong correlation with human creativity judgments, providing a reliable automated metric for P-Creativity evaluation at scale. Applying this pipeline to agent trajectories, we find: (1) all agents exhibit declining P-Creativity as they transition from exploration to exploitation; (2) LLMs exhibit greater H-Creativity than medal-winning humans, yet achieve lower performance. Our findings reveal that current agents can explore novel regions of the solution space but lack the capacity to convert this novelty into improved task performance.

Noisy labels remain a critical challenge for training deep neural networks, since memorizing incorrect labels degrades generalization. Once noisy samples are identified after training, the standard solution is to retrain the model from scratch on the cleaned dataset, which is increasingly expensive as datasets and models grow. Machine Unlearning (MU) has recently emerged as a computationally efficient alternative, but the relative effectiveness of different MU strategies for noisy-label correction remains poorly understood. In this work, we conduct a comparative empirical study of five MU methods (NegGrad, Fine-Tuning (FT), Random Labeling (RL), SalUn, and MUNBa) across symmetric, asymmetric, instance-dependent, and open-set noise on CIFAR-10, CIFAR-100, and the real-world noisy dataset Food-101N. Our central finding is that the appropriate unlearning strategy is conditioned on the noise structure. Simple FT is a strong baseline across most closed-set scenarios; RL and SalUn are the most consistently robust methods and, under instance-dependent noise, approach retraining accuracy at a fraction of the computational cost; MUNBa shows advantages mainly under extreme symmetric noise. Under open-set noise, in contrast, we show that retraining on the cleaned subset degrades accuracy relative to the noisy baseline, so approximating the retrained model is not an adequate objective in this regime. On Food-101N, all MU methods remain competitive and achieve accuracies close to retraining despite reducing runtime by an order of magnitude. These findings provide practical guidelines for selecting MU strategies for post-training noisy-label correction.

Model comparison increasingly relies on large collections of publicly reported benchmark scores, yet common aggregation strategies trade off evidence coverage against control over capability weighting. Manually curated suites leave potentially informative evaluations unused, while uniform averaging retains them but gives greater influence to capabilities that happen to be benchmarked more densely. We introduce Balance of Benchmarks (BoB), a framework that retains eligible benchmark evidence while adapting its influence for task-conditioned model comparison using only public aggregate scores. BoB combines semantic density weighting, score equating across benchmarks of different difficulty, and task-relevant residual pooling. We evaluate it on 605 configurations across 14 Artificial Analysis benchmarks and on WildScores, a collection of 148 developer-reported benchmarks evaluated with held-out source-lineage families. On WildScores, BoB-Support raises family-mean Spearman correlation from 0.764 under uniform standardized averaging to 0.823, reduces MAE from 6.19 to 5.10 normalized score points, and increases three-model shortlist hit rate from 65.3% to 72.6%. BoB-Constant reaches a Spearman correlation of 0.831 and a hit rate of 74.6%. Separately, density weighting reduces average ranking changes when benchmarks are repeated, including as paraphrased copies. BoB-Support also reduces retrospective three-model shortlist regret from 2.08 to 1.67 normalized score points. BoB makes benchmark inclusion, redundancy, and task relevance explicit and testable measurement choices, allowing existing benchmark evidence to be used more fully while moderating the influence of benchmark proliferation.

We construct Fortin operators for discontinuous Petrov-Galerkin (DPG) discretizations of the advection equation with a piecewise constant divergence-free advection vector $\beta$, on simplicial meshes of any spatial dimension $N$ and for any polynomial degree $k \ge 1$ of the trial space. A minimal test space is built on each element from facet and interior bubbles and later augmented. The Fortin operator is shown to be bounded, uniformly over shape-regular mesh families, in the natural $\beta$-weighted broken test graph norm built on $L_q$ for every $1 < q < \infty$, where $q$ is the exponent conjugate to the trial exponent $p$. A non-characteristic facet condition is assumed when $k \ge 2$, while the lowest-order case requires no such condition and admits characteristic facets. As applications we prove that the practical fully discrete residual minimization method is quasioptimal in the DPG energy norm for every $1 < p < \infty$, with a quasioptimality constant governed solely by the Fortin operator, that its computable residual is a globally reliable and efficient a posteriori error estimator, and that augmenting the test space and changing the test norm improves the results in the lowest-order case.

Large language model agents place outputs from external skills into their execution context, allowing attacker-controlled data to influence later privileged actions. Existing defenses mainly classify untrusted content or authorize proposed operations. They do not directly address how an agent's future authority should change once untrusted data enters its state. We present SkillGuard, a harness-level enforcement layer that treats this event as contamination and restricts future capabilities to disconnect the resulting state from deployer-defined forbidden states. Given sound skill summaries and policies, SkillGuard represents security-relevant transitions with a Skill Impact Graph, specifies admissible control over skill parameters via steerability signatures, and mediates invocations with an inline reference monitor. Following contamination, it computes weighted capability restrictions using binary, fractional, or fractional-flow strategies without auxiliary language-model inference. We evaluate SkillGuard on four AgentDojo suites with two backend LLMs, Gemini 2.5 Flash and Llama3.3-70B, against an LLM-only No Defense baseline and three defenses at different system layers: Spotlighting, CaMeL, and AttriGuard. We construct a compositional attack benchmark in which each attack combines observations individually insufficient to induce target violation and evaluate the same baselines on it. Under AgentDojo's Tool Knowledge attacks, SkillGuard eliminates attack success on three of four suites for both backends and reduces it to 4.8% and 14.3% on Slack. Against compositional attacks, it outperforms every baseline on Llama and matches the strongest baseline on Gemini at higher benign utility. Fractional-flow restriction preserves substantially more capabilities than binary restriction at the same attack success rate. Across both settings, SkillGuard adds no model calls or token overhead.

Missing data, measurement error, and population heterogeneity are pervasive challenges in analyzing data arising from modern observational studies and machine learning applications. Although these problems frequently coexist and interact, they are often treated separately in existing works. We propose a unified probabilistic framework that jointly addresses these issues utilizing deep latent variable representation. The proposed method integrates a novel hierarchical tree-routed variational autoencoder with pattern-aware latent representations and calibration-based denoising. The framework accommodates missing data mechanisms, including MCAR, MAR, and MNAR, while simultaneously learning subgroup-specific and globally shared latent structure. The introduced reconvergent routing mechanism enables selective parameters to be shared across related subpopulations, which offers flexibility as well as improved statistical efficiency. Simulation studies demonstrate substantial improvements over existing deep generative imputation approaches under complex heterogeneous missingness and measurement-error settings. The proposed framework provides a principled approach for learning from noisy and incomplete data in modern healthcare and other high-dimensional applications.

System-level mobile agents are evolving from fixed scripts into adaptive systems that continuously observe interfaces, reason, and adjust their actions, allowing automated attacks to navigate dynamic UIs and complete complex tasks. Existing applications detect automation using touch trajectories, action timing, and the physical coupling between touch and inertial measurement unit (IMU) signals. However, a privileged system-level agent executor can control both touchscreen input and application-visible sensor delivery, enabling it to jointly generate time-aligned touch and six-axis IMU signals and evade these defenses. We present ActReal, a physical-action attack framework for system-level mobile agents. ActReal converts semantic agent actions into task-valid touch and IMU events using genuine-trajectory adaptation and physics-guided IMU generation. ActReal achieves a mean event-level attack success rate of 77.5%; even when detectors jointly observe touch and IMU, its attack success rate remains 71.1%.

Given a finite set $A$ of natural numbers whose greatest common divisor is one, the Frobenius number $g(A)$ is the largest integer that is not a non-negative integer combination of the numbers in $A$. In a 2016 preprint, Matsubara states that given $A$ and $k$, deciding if $g(A) \geq k$ is $\Sigma_2^P$-complete. A decade has passed since without peer-reviewed publication of this result. At the same time, the community has found it difficult to verify this result. In this note, we give a write-up of the completeness proof based on Matsubara (2016).

Self-evolving reasoning frameworks train a Challenger to generate questions exposing a Solver's weaknesses, creating adaptive curricula without human data. However, existing approaches use a single solver's sampling uncertainty as the Challenger's reward. This creates a fundamental bottleneck: as the solver grows confident on the Challenger's question distribution, all sampled answers converge identically, collapsing the reward to zero and starving the Challenger of learning signal. Critically, this single-model reward cannot distinguish genuinely easy questions from those that merely align with one solver's learned biases. We propose a multi-solver disagreement reward using a heterogeneous ensemble varying in model capacity and sampling temperature. A normalized Shannon entropy over the ensemble's per-question plurality answers explicitly rewards questions where solvers produce conflicting solutions—capturing difficulty as inter-model divergence rather than intra-model sampling variance. This richer gradient enables the Challenger to discover questions targeting true capability boundaries, producing a curriculum that forces downstream Solvers to develop robust reasoning strategies generalizing across problem types. Our approach is a drop-in reward function replacement requiring no framework modifications or additional data. Experiments with Qwen3-4B show that Solvers trained on disagreement-Challenger questions achieve +1.34 points average improvement on competition-math benchmarks (MATH-500, AMC, Olympiad), suggesting that multi-solver disagreement provides a complementary and scalable signal for curriculum generation in self-play reasoning systems.

Two-stage vision pipelines often place an enhancement network before a task network, on the assumption that a cleaner input produces a better output. We evaluate this in a firefighter assistance pipeline, where a dehazer precedes an edge detector that renders smoke-filled rooms as structural outlines. Both were designed for a Raspberry Pi 4, at 355K and 23K parameters, and quantized to UINT8 via TensorFlow Lite. The float dehazer reaches 18.60 dB peak signal-to-noise ratio (PSNR) on held-out real smoke against 13.60 dB unprocessed and 17.08 dB for an AOD-Net trained on the same data, and the edge detector reaches an F-measure at optimal dataset scale (ODS) of 0.738, outperforming an optimized Canny's result of 0.692. Dehazing improves edge extraction under dense smoke but degrades it on clear and lightly hazed frames, where the dehazer discards more detail than the haze obscures. We therefore run the dehazer only when a dark channel haze estimate exceeds a threshold, a 10.1 ms test that lets the pipeline save 469.6 ms on the dehazing stage. Averaged over four haze levels, gating is more accurate than either fixed decision, at 0.675 mean ODS against 0.664 for always dehazing and 0.630 for never dehazing. It reduces the mean per-frame time on the Raspberry Pi from 569 ms to 321 ms, and on clear frames increases the frame rate fivefold, from 1.8 to 9 frames per second.

Large language models (LLMs) are increasingly used to simulate human behavior but frequently fail to exhibit realistic cognitive constraints, suffering from a "superhuman bias." Using a dataset of over 71,000 reading comprehension responses from 2,359 primary-school students (grades 4–6), we demonstrate that standard persona prompting yields near-perfect, deterministic performance, failing to capture the natural variance of developing readers. To address this, we introduce the Cognitively Bounded User Simulator (CBUS), an architectural framework that explicitly models the restricted working memory of young readers through an episodic bottleneck. Within this framework, we formalize two distinct test-taking strategies to emulate different reading behaviors. Our evaluation shows that explicitly modeling cognitive bounds significantly narrows the simulation gap across multiple LLM backbones, demonstrating that enforcing architectural constraints is more effective for high-fidelity simulation than simply scaling raw model capabilities.