Papers

Large Language Model (LLM)-based counseling agents can generate fluent and supportive responses, but they often lack the structured, goal-directed progression required to conduct a coherent therapeutic session. We present DeepSAGE (Strategic AI Guidance Engine), a hybrid LLM–Deep Reinforcement Learning (DRL) framework for stage-aware counseling dialogue grounded in the first session of Cognitive Behavioral Therapy (CBT). DeepSAGE represents the session as eleven stages with explicit therapeutic objectives, with an external controller determines stage completion and the DRL model selects therapeutic intentions that guide LLM response generation. We evaluate DeepSAGE against six retrieval-, prompting-, stage-, and policy-based alternatives. DeepSAGE elicits higher simulated client engagement and openness and achieves the strongest balance of stage-goal completion and dialogue efficiency among stage-structured systems. Domain expert review further indicates that the generated conversations exhibit broadly plausible emotional trajectories and recognizable CBT processes. Because the evaluation relies primarily on simulated clients and model-based metrics, these findings demonstrate comparative dialogue-control improvements rather than clinical effectiveness. These results suggest that combining stage-structured dialogue with learned strategy selection is a promising approach for AI counseling, though clinical effectiveness, safety, and real-world utility require further human evaluation.

Interpretability research increasingly asks when concepts emerge during training and whether linear probes recover real structure, but in language models these claims are hard to validate because language offers little ground-truth ordering of concepts or relationships among them. We propose the use of astronomical ground truth through AstroPT, a transformer trained on millions of galaxy images, as a calibration testbed. AstroPT is an LLM-like model trained within a domain where the difficulty ordering of concepts and the relations among them are known in advance. Probing frozen representations across checkpoints, layers, model sizes, and objective choices, we find that galaxy properties emerge in a fixed order that tracks their known difficulty—quantities written almost directly into the pixels (band magnitude) become decodable early in training and shallow in the network, while multiband/spectra based and inferred quantities (such as redshift and specific star formation rate) emerge later and deeper. This order is invariant to our tested training objectives, and scales in magnitude but not in sequence with capacity. Our linear probe directions further recover the known physical structure among galaxy properties. Our findings suggest that astronomy offers a controlled sandbox for calibrating mechanistic interpretability methods we otherwise apply to LLMs blind.

The rapid evolution of hybrid large language models (LLMs), which interleave grouped-query-attention (GQA), state-space model (SSM), and Mixture-of-Experts (MoE) layers, introduces two fundamental challenges for near-memory processing (NMP) architectures. First, the Technology Wall: the conventional 6F^2 DRAM cell is approaching its physical scaling limits at 10nm-class nodes, making it difficult to meet the memory capacity demands of MoE models with hundreds of experts. Second, the Architecture Wall: existing NMP designs target narrow arithmetic intensity (Op/B) ranges and cannot efficiently support the heterogeneous compute characteristics of hybrid LLMs, spanning memory-bound SSM layers, compute-intensive GQA layers, and large Op/B variations across experts. We propose NOVA, a technology-architecture co-designed NMP system that overcomes both walls. On the technology side, NOVA combines a 4F^2 vertical channel transistor (VCT) DRAM cell with a peri-over-cell (POC) structure to achieve approximately 2x memory density at iso-area over conventional 6F^2-based DRAM, enabling continued scaling into sub-10nm nodes. On the architecture side, NOVA repurposes the POC peripheral-die (peri-die) to host processing units, forming a 2-tier NMP architecture: Tier-1 (peri-die NMP) for low-to-mid Op/B operations, and Tier-2 (base-die NMP) for mid-to-high Op/B operations. Parallel execution across tiers supports diverse compute patterns for hybrid LLMs, maximizing inference performance. Evaluated on state-of-the-art hybrid and MoE LLMs including Nemotron3-Nano, Nemotron3-Super, Falcon-H1R, and Qwen3, NOVA achieves on average 4.5x higher throughput, 69.8% lower end-to-end latency, and 5x better energy efficiency over a GPU baseline, with only 3.9% area overhead and no loss in memory capacity.

Agent skills, structured artifacts distilled from interaction trajectories and dynamically reused from skill banks, have become a central mechanism for enabling large language model (LLM)-based self-evolving agents to learn from past experience. Yet existing work has largely focused on the operational aspects of skills, such as acquisition, evolution, and retrieval, while leaving a more fundamental reliability question unresolved: Do accumulated skills in an agent's skill bank actually make positive mechanistic contributions? We investigate this question through systematic skill-bank audits across alternative bank compositions and deployment domains, measuring the resulting changes in agent behavior. These audits reveal two recurring reliability failures: coalition pollution, where bank-level gains conceal negative coalition-level skill contributions, and cross-domain utility reversal, where source-beneficial skills reverse their effects after transfer. These findings motivate two reliability interventions: coalition-aware skill selection during skill accumulation and label-free skill masking after transfer. Coalition-Aware Skill Selection (CASS) selects more reliable candidate skills for the current bank using sampled Shapley marginals. Unsupervised Skill-Masked Coalition Optimizer (u-SMCO) masks transferred skills whose exclusion improves retrieval quality on unlabeled target-domain data. Agentic experiments on LoCoMo, LongMemEval, HotpotQA, and ALFWorld show that CASS and u-SMCO consistently improve task performance and cross-domain generalization over strong skill-based self-evolving agent baselines. Beyond accuracy, coalition-conditioned reliability modeling reduces sensitivity to noisy outcome-reward fluctuations during reinforcement learning and exposes the limits of isolation-based skill evaluation.

Financial fraud detection relies heavily on centralized machine learning models. This creates serious data privacy risks. Federated Learning (FL) decentralizes data processing, but financial regulations still require models to be transparent. This means using Explainable AI (XAI) tools such as TreeSHAP. Recent cybersecurity research shows a problem with this approach. Sharing high-fidelity SHAP explanations exposes the federated network to Membership Inference Attacks (MIAs). This dissertation proposes and evaluates DP-FedSHAP. It is a new architecture that applies client-level differential privacy only to post-hoc TreeSHAP vectors. It is compared against a Weight-Level DP baseline, which perturbs the trained model directly instead. Using the highly imbalanced IEEE-CIS Fraud Detection dataset, this study measures the trade-off between explanation fidelity, privacy preservation, and the model's Area Under the Precision-Recall Curve (AUPRC).

Topology optimization, using both physic-based approaches and deep learning surrogates, serves as a cornerstone for generative design agents in cyber-manufacturing systems. While deep learning surrogates have gained widespread adoption due to their speed in online design generation, this work demonstrates their vulnerability under input perturbations. In this work, we present a mechanics-grounded reliability evaluation framework that formulates an adversarial agent targeting the generative design models. We investigate a strictly non-intrusive threat model where bounded perturbations are introduced exclusively to the initial-density channel, while physical boundary conditions, compliance-gradient channels, network architectures, and solver routines remain intact. Evaluating surrogate models across U-Net, convolutional, and generative architectures with varying physics-gradient conditioning depths demonstrates that bounded initialization noise can cause catastrophic mechanical failure, increasing compliance by multiple orders of magnitude through severed load paths and disconnected supports. Furthermore, we discover that incorporating richer physics-gradient conditioning in the deep learning surrogates does not guarantee monotonic robustness across surrogate families. Finally, physics-in-the-loop recovery demonstrates that initializing the classical SIMP optimizer with perturbed topologies mitigates design performance degradation, having a high probability of restoring compliance to near-baseline levels across tested instances. These findings demonstrate that learned surrogates should serve as physics-verified initializers instead of replacing physics-based solvers entirely in a resilient cyber-manufacturing system. Moreover, the proposed adversarial agent provides a foundation for future training generative design agents robust against noise and targeted perturbations.

The rapid evolution of machine learning workloads has fundamentally transformed GPU hardware, driving architectures toward Multi-Chip Module (MCM) topologies, asynchronous execution primitives, and persistent, multi-phase kernel behaviors. Despite these shifts, cycle-level simulation infrastructure has lagged behind, lacking the native capability to model the physical non-uniformity of modern GPUs alongside the massive scale of state-of-the-art AI workloads. To bridge this gap, we present a cycle-level simulation framework designed to accurately model modern GPU generations, including Ampere, Hopper, and Blackwell. Rigorously validated against physical silicon, the simulator achieves a 99% Pearson correlation coefficient and a 13.4% mean absolute cycle error on the H100 GPU. Utilizing this infrastructure, we conduct architectural case studies to evaluate emerging design trajectories, including chiplet topology scaling, expanded SRAM capacity and bandwidth, and inter-GPU prefetching strategies.

Subsampling is effective in tackling computational challenges for massive data with rare events. Overly aggressive subsampling may adversely affect estimation efficiency, and optimal subsampling is essential to mitigate the information loss. However, existing optimal subsampling probabilities depend on data scales, and some scaling transformations may result in inefficient subsamples. This problem is more significant when there are inactive features, because their influence on the subsampling probabilities can be arbitrarily magnified by inappropriate scaling transformations. We tackle this challenge and introduce a scale-invariant optimal subsampling function in the context of sparse models, where inactive features are commonly assumed. Instead of focusing on estimating model parameters, we define an optimal subsampling function to minimize the prediction error, using adaptive lasso to outline the estimation procedure and study its theoretical guarantee. We first introduce the adaptive lasso estimator for rare-events data and establish its oracle properties, thereby validating the use of subsampling. Then we derive a scale-invariant optimal subsampling function that minimizes the prediction error of the inverse probability weighted (IPW) adaptive lasso. Finally, we present an estimator based on the maximum sampled conditional likelihood (MSCL) to further improve the estimation efficiency. We conduct numerical experiments using both simulated and real-world data sets to demonstrate the performance of the proposed methods.

We develop a new framework for flexible, nonlinear, and interpretable off-policy evaluation for infinite-horizon reinforcement learning. To handle large state spaces and support transparent decision-making, we model the Q-function using a nonlinear function class with a sparse additive structure. We derive high-probability finite-sample error bounds for estimating the value function of a target policy and show that the bounds depend only logarithmically on the ambient dimension $d$, thereby alleviating the curse of dimensionality. In contrast to most existing theory for off-policy evaluation, which typically assumes access to many trajectories, our analysis guarantees accurate value estimation when either the number of trajectories or the time horizon is sufficiently large. In addition, we propose a group-sparsity-based feature screening procedure that identifies, with high probability, a reduced feature set containing all relevant covariates. Numerical experiments demonstrate the effectiveness of the proposed approach.

Tabular foundation models (TFMs) have recently emerged as a promising paradigm for machine learning on tabular data, offering the ability to generalize across datasets without task-specific training. Since many machine learning datasets can be represented as tables, this raises the question: does TFM capability extend beyond tasks traditionally regarded as tabular? We address this question by using TabPFN v3 on three non-tabular classification problems: handwritten digit recognition on MNIST, language identification of French and German words, and image classification on Tiny ImageNet. In each case, the original data are represented as rows of a table and classification is formulated as prediction of a missing label. We evaluate performance as a function of the number of context samples provided to the pretrained model, with no additional training or fine-tuning. Despite having no explicit access to the spatial or sequential structure characterizing the data, TabPFN v3 in some cases achieves accuracies comparable with that of models or methods geared specifically toward the corresponding tasks.

Autoencoder (AE)-based federated learning (FL) is attractive for anomaly detection when clients have limited local data. However, conventional FL exchanges AE parameters or gradients, incurring substantial communication overhead and potentially exposing input training data information, since AEs are explicitly optimized to reconstruct their inputs. We introduce Global Centroid Alignment (GCA), a latent-code-mediated FL protocol that coordinates clients without transmitting AE parameters or gradients. In each round, (1) clients first train their local AEs using a reconstruction update and upload a small subset of encoder latent codes to the FL server. (2) The server pools these codes, fits a clustering model, and broadcasts only global latent centroids and their support counts. (3) Each client then updates its encoder by aligning its local latent codes with the nearest centroid using inverse-count weighting to emphasize globally underrepresented patterns. Steps (1)–(3) repeat over communication rounds. Because GCA exchanges only sampled latent codes and centroid statistics, its communication cost depends on latent dimensionality and the numbers of uploaded codes and returned centroids rather than on AE model size. Across five tabular and two vision benchmarks, GCA yields higher reconstruction error under a server-side client data extraction attack in all 21 comparisons and clearly lower cosine similarity in 20 of 21 comparisons with FedAvg, FedProx, and FedNova, showing its ability to protect training data. It even improves test accuracy over FedAvg by up to $5.76\%$. GCA achieves extraction defense comparable to DP-FedAvg, remains effective when DP-FedAvg does not reduce target resemblance, and lowers per-round communication by up to $99.15\%$.

Gradient random walk (GRW) methods represent a solution's spatial derivative by weighted particles and recover the solution by cumulative summation. We examine their accuracy for the one-dimensional heat, scalar cubic reaction-diffusion, and Burgers' equations. Thirty-seed ensembles, paired reconstructions, and deterministic controls identify the sources of two apparent accuracy limits. For the heat equation, comparing cumulative bin sums at bin centers introduces a first-order alignment error, which the exact reflected-particle distribution predicts and bin-edge comparison removes. An identity for the expected squared error predicts the remaining floor and the particle count for a target root-mean-square accuracy. With unequal weights, a count chosen in advance gives an error within 1% of its prediction. Reaction-front position, speed, and shape errors decrease under refinement. For Burgers' equation, exact transformed boundary data reduce the deterministic Cole-Hopf error about 490-fold, while fixed-endpoint particle runs approach their deterministic reference.

Vision-language-action policies inherit both capabilities and input representations from pretrained vision-language models, showing great potential for robotic manipulation across diverse industrial settings. As these policies increasingly use interaction history, organizing the representation of historical observations determines how experiences enter temporal context and how relationships across time are modeled, which is a generally ignored challenge in previous works. In this work, we believe solving this challenge requires an architectural reconstruction and propose WorldToken. WorldToken encodes each timestep's observations into one world token, processes the resulting history with a causal Transformer, and generates action chunks with a diffusion action head. Unified token enables long horizon tasks while relieving the memory requirement of the temporal backbone, hence improving performance. This design also offers high interpretability and allows advances in language modeling, such as pre-training and scaling, to be transferred to robot interaction policies. In RoboCasa experiments, WorldToken successfully handles most tasks with 85M parameters and achieves 59.4% mean closed-loop success close to $\pi_{0.5}$ with 3.35B parameters. On the memory benchmark RMBench Blocks Ranking, WorldToken can reach the context of two minutes and achieves a success rate of 95%. In addition, holdout action RMSE is well described by power-law fits, and its closed-loop success rate improves consistently with increasing training data size in a study involving approximately 350,000 closed-loop evaluation episodes across 50 trained policies on RoboCasa, showing its scaling potential. We also conduct experiments to analyze information preservation and history use in WorldToken, providing empirical grounding for future work.

External hand forces are important inputs to biomechanical analyses of occupational physical exposure and injury risk, yet continuous force measurements during manual material handling (MMH) typically requires instrumented objects or specialized sensing. We evaluated a vision-language model (VLM)-based pipeline that combines task-specific textual cues, visual representations, and known box mass to estimate dynamic, triaxial, bilateral external hand forces from RGB video. Thirty-five healthy young adults performed five MMH tasks involving lifting, carrying, pushing, and pulling with box masses of 6, 9, and 12 kg. The pipeline used text-guided localization of participant and handled-object regions of interest (ROIs), pretrained vision-transformer feature extraction, and transformer-based temporal regression. Performance was evaluated using leave-one-subject-out validation across seven camera-view conditions (three single-view and four multi-view conditions) and four ROI strategies. Overall, root mean square error was ~4.7-5.6 N for the horizontal and mediolateral force components and ~10.6-11.0 N for the vertical component. Including the handled object as a second ROI generally improved force estimation, with some of the largest benefits under single-camera conditions, whereas pixel-level segmentation provided little additional improvement. Multi-camera capture provided the clearest benefit for peak-force estimation, particularly for the vertical component, whereas differences in overall frame-level error among camera configurations were comparatively modest. These findings demonstrate the feasibility of estimating continuous, bilateral, directional hand-force estimates from RGB video and known load mass without requiring sensors on the worker or handled objects as model inputs, supporting the development of more scalable occupational physical exposure and risk assessments.

Two-stage neuro-symbolic architectures provide an elegant paradigm for visual problem solving by cleanly separating connectionist perception of predefined symbols from possibly later defined relational reasoning thereon. However, anchoring high-level predicates into visual frames typically necessitates annotations that are expensive to acquire. In this work, we introduce the Dynamic Orthogonal Concept Bottleneck (D-OCB), an object-centric slot- VAE framework designed to extract human-aligned symbolic predicates under extremely weak supervision. D-OCB eliminates the arduous manual tuning of loss-balancing coef- ficients by dynamically learning optimal hyperparameter allocations during training. To infuse prior knowledge on independence of concept categories, in addition to standard re- construction self-supervision we penalize correlation across concept subspaces. Crucially, to combat the instability of very low supervision regimes, D-OCB incorporates a dynamic di- mensionality allocation mechanism; this adaptive formulation allows well-represented con- cepts to yield latent dimensions to underperforming concepts that are lagging behind, effectively preventing representation collapse and significantly improving overall concept accuracy. Through an extensive empirical evaluation, we demonstrate that our framework achieves high concept alignment and downstream visual reasoning accuracy using minimal label budgets, matching or outperforming end-to-end paradigms.

Clinical records contain rich evidence about patient state, but converting that evidence into reliable, structured knowledge graphs remains difficult because extraction errors, ontology mismatch, missing relations, and temporal ambiguity can propagate into downstream systems. We propose a clinical knowledge graph construction and refinement framework that combines multi-agent relation proposal, ontology-aware normalization, deterministic evidence scoring, and JEPA-based latent refinement. Rather than treating a clinical knowledge graph as a static extraction artifact, we treat it as a predictive patient-state representation. For each admission, the system constructs an evidence-scored graph from structured MIMIC-IV records and inferred clinical cross-links, then learns to recover held-out clinical relations from the observed graph context. We evaluate the refiner with leakage-free leave-one-out edge recovery (MRR and Hits@k) and held-out batch-mask evaluation (AUC and MRR). To isolate the contribution of discharge-note context, we compare a note-embedding-free configuration with a note-augmented configuration that injects real discharge-note representations only into note-grounded entities. Under the same cohort and evaluation protocol, entity-grounded note injection improves overall leave-one-out MRR by 31% relative improvement.

Large-scale population surveys are essential for generating robust social and scientific insights, yet they face significant challenges, including declining response rates, increasing data collection costs, long delays between data collection and data provision, and the risk of nonresponse bias. Advances in artificial intelligence (AI) have opened up new opportunities for AI-supported survey infrastructures where the goal is to overcome these challenges without limiting the data quality. A promising AI-enabled survey infrastructure for which we build a first pilot is a hybrid panel. A hybrid panel is a longitudinal AI-enabled survey which allows to iteratively improve the alignment between large language models (LLMs) and the population they aim to simulate and use the errors to inform the design and implementation of the next survey wave (e.g., inform the participant recruitment, assignment of questions to participants). It incorporates both human participants and LLMs as fundamental elements of its design. In this research note, we introduce the concept of a hybrid panel by providing a definition and outlining an overarching framework, spanning data collection to data validation. We detail results from a first pilot study to illustrate (open) challenges that we identify for hybrid panels.

Long-horizon GUI agents can retain complete action histories as compact text, but only a few historical screenshots fit in active context. We formulate this as budgeted fidelity restoration: every event remains summarized, while a fixed budget $B$ determines which events regain their archived screenshots. Recent-$B$ assigns all visual slots to the latest events. CausalCache instead scores the complete history and swaps in an older event only when its predicted utility exceeds that of a recent event. A history-gated key/value adapter modifies only restored history-image tokens and is exactly bypassed when no history image is active, preserving current-screen processing. The adapter and selector are trained with matched-budget interventions on desktop trajectories and evaluated zero-shot on mobile. On OSWorld-Verified, activating historical screenshots improves success by about $13$ percentage points over summary-only memory. Under the official $15$-step limit, CausalCache and Recent-$4$ are statistically indistinguishable; in a $30$-step diagnostic, CausalCache achieves $46.7\%$ success versus $42.4\%$ ($+4.3$ points). Zero-shot on $117$ MobileWorld tasks, CausalCache improves over Recent-$4$ from $30.2\%$ to $36.8\%$. The gain is concentrated on a pre-defined cross-app memory-candidate split ($30.6\%$ vs. $19.4\%$, $+11.2$ points), while single-app controls show no detectable difference ($43.6\%$ vs. $42.4\%$). These results show that selecting which past events regain pixels is more effective than spending a fixed visual budget entirely on recency.

Graph neural networks (GNNs) have emerged as a powerful framework for learning from graph-structured data. However, their efficient training remains challenging, particularly in distributed computing environments. This challenge arises from the use of message passing, which couples all graph nodes, leading to expensive optimization steps, high memory requirements, and substantial communication overhead. To alleviate these limitations, we propose a novel domain-decomposition (DD) variant of AG2m, an AdaGrad method enhanced with second-order curvature information and momentum, denoted by DD-AG2m. The proposed DD-AG2m alternates between AG2m optimization on the original (global) graph and AG2m optimization on the partitioned graphs. To incorporate global information at reduced cost, we further introduce a two-level variant (2DD-AG2m) that performs global optimization steps on a coarse graph obtained by randomly subsampling nodes within each subdomain. Numerical experiments spanning graph classification, node-level regression, and spatiotemporal forecasting tasks demonstrate that the proposed DD methods reduce the computational cost required to achieve the same predictive performance by a factor of 4-8. Moreover, for the fixed computational cost, they improve the predictive performance of GNNs by up to 22% compared with the baseline AG2m.

In today's digital landscape, high-efficiency video coding (H.265/HEVC) has emerged as the most widely used video coding standard, employing selective encryption schemes to protect the privacy of video content while maintaining efficient compression performance. However, existing coefficient scrambling methods impose a significant computational load, leading to increased bit rate overhead due to encryption, longer execution times, and insufficient safety measures. To address these issues, a new coefficient scrambling scheme based on chaotic maps is proposed. This approach leverages the pseudorandomness, ergodicity, and sensitivity to initial conditions inherent in chaotic maps to generate highly unpredictable coefficient distributions, thereby strengthening security while preserving low complexity. Unlike conventional scrambling, chaotic maps ensure minimal correlation between encrypted coefficients, enhancing resistance against statistical and differential attacks. Additionally, the scrambling conditions are specifically designed to minimize the impact on the bit rate overhead. Furthermore, when combined with syntax element encryption (SEC), which includes motion vector difference (MVD), quantized transform coefficients (QTC), and luma intraprediction mode (Luma IPM), this method effectively distorts video content. The proposed scheme operates synchronously with slices, ensuring that the decryption of video content remains intact even if some slices are lost. Additionally, a random sequence generated by AES-CTR is incorporated with the H.265 encoded stream to protect against chosen-plaintext attacks.

Ensuring reliability, safety, and economic efficiency in airline operations requires maintenance and fleet scheduling strategies that explicitly account for uncertainty in Remaining Useful Life (RUL) predictions. However, the integration of prognostic uncertainty into operational decision-making remains a major challenge. In practice, tail assignment (TA) and maintenance scheduling (MS) are typically optimized separately or sequentially, thereby limiting the effective use of predictive health information despite their strong interdependencies. This paper proposes a unified optimisation framework that jointly integrates TA, MS, and predictive maintenance (PdM) under RUL with confidence intervals. The problem is formulated as a stochastic mixed-integer linear program, and a scalable solution approach is developed by embedding a neural network surrogate to approximate expected disruption costs resulting from RUL uncertainty. The proposed framework is evaluated using operational scenarios derived from real-world airline data. Results show that explicitly incorporating prognostic uncertainty in a joint planning model reduces operational risk, i.e., downstream disruption costs and flight cancellations, compared to deterministic and sequential approaches, at the expense of moderate increases in planning cost. These findings highlight the value of tightly coupling predictive maintenance with operational planning and demonstrate the potential of surrogate-assisted stochastic optimisation for scalable, uncertainty-aware airline decision-making.

Multi-agent large language model (LLM) systems are designed to improve reasoning by decomposing tasks across multiple agents with specialized functions, but the presence of multiple agents does not inherently guarantee coherent reasoning or outputs that align with task objectives. This paper introduces a quantitative ethnographic (QE) approach for diagnosing and redesigning multi-agent LLM systems based on the discourse produced through agent interactions. We test this approach using automated essay scoring as an example context, applying Epistemic Network Analysis (ENA) to model a five-agent multi-agent debate system and examine differences between debates that produced correct versus incorrect scoring decisions. Results show that, in the initial system, correct scoring decisions were characterized by rubric-grounded justification, agreement, and elaboration. Incorrect scoring decisions, in contrast, were characterized by extended proposition-challenge-response exchanges that were less consistently tied to rubric criteria. We then used the findings to revise the agents' prompts. The revised system improved exact scoring accuracy from 27.78% to 40.28% and shifted the discourse of incorrect debates toward the rubric-grounded pattern of correct ones, making the two nearly indistinguishable. Based on these results, we argue that QE can support a diagnostic-to-redesign loop for AI reasoning by tracing how patterns of agent interaction relate to system performance, informing prompt redesign, and evaluating whether those redesigns change both outcomes and interaction patterns.

In adaptive control, parametric uncertainties in linear-in-parameter form consist of unknown parameters and known regressor signals. Convergence of the unknown parameters to their ideal values requires the regressor to satisfy a persistent excitation (PE) condition, which depends on future data and is therefore infeasible to guarantee online. Memory-based parameter update laws address this by enabling ideal parameter convergence under the online-verifiable finite excitation (FE) condition. In this paper, a new algorithm is proposed to construct a memory term via the Modified Gram-Schmidt orthogonalization procedure for a class of multi-input multi-output nonlinear systems with an unknown diagonal control effectiveness matrix and bounded nonparametric uncertainties. Under the finite excitation condition, the constructed memory term yields an identity coefficient matrix in the parameter estimation error dynamics. The identity coefficient matrix eliminates the need for time-varying adaptation gains, enables an explicit ultimate bound on the parameter estimation error, and preserves the structure of the nonparametric uncertainty bound under the memory term. Building on this, a combined adaptation law is developed for controller gain estimation under FE. The closed-loop tracking and estimation errors are shown to decay exponentially to a neighborhood of the origin, characterized by an explicit ultimate bound, with a decay rate that depends solely on user-defined gains and system constants, independent of the level of regressor excitation. This removes the dependence of the convergence rate on the level of regressor excitation, a key limitation of existing approaches such as concurrent learning, memory regressor extension, and DREM.

We consider the compression of kernel matrices on the unit interval $[0,1]$ by interpolets that have sufficiently many vanishing moments. We define a compression rule which discards most matrix coefficients without compromising the accuracy offered by the underlying discretization. Since interpolets can be scaled such that the compressed kernel matrices are well conditioned, we derive a fully discrete scheme that solves a kernel interpolation problem under consideration in linear overall complexity. We finally generalize this approach to the unit $n$-cube $[0,1]^n$ by means of the sparse grid combination technique. Numerical experiments are carried out to validate the theoretical findings.

Rubrics aim to make language-model evaluation transparent by decomposing response quality into interpretable criteria. However, natural-language rubrics are often ambiguous, require LLM judges, and typically assume criteria aggregated through linear weighted sums, limiting their ability to capture dependencies, alternatives, penalties, and override conditions. We propose ExecRubrics, a framework for representing rubrics as compact executable programs. ExecRubrics encodes evaluation logic as verifiable Python scoring functions, giving natural-language rubric intent an operational semantics: a fixed decision procedure that can be inspected, executed, and edited. On three long-form response benchmarks – HealthBench, HelpSteer, and ArgQuality – we show that ExecRubrics can recover substantial preference signal without an LLM judge at evaluation time. On ArgQuality and HelpSteer, the strongest executable variants are within 1.1 and 4 percentage points, respectively, of the direct GPT-5.5 agentic baseline. Executable rubrics are also considerably faster, achieving a 192x average speedup. We show that incorporating external logic and resources from text processing libraries such as NLTK and spaCy can further improve preference accuracy. Our results suggest a novel way of approaching automated evaluation, by offering a faster, more explainable, and less ambiguous alternative to black-box rubric evals, particularly in high-stakes domains such as healthcare and banking where precision and auditability are critical.

Existing LLM unlearning methods struggle with robustness: unbounded forget losses degrade model coherence, fixed-weight balancing cannot adapt as retain difficulty shifts mid-training, and methods that work on one benchmark falter under scaling or repeated application. We propose BLADE, a constrained bilevel framework whose three mechanisms give smooth, predictable control over the optimization landscape: a clamped-entropy forget loss whose gradient is exactly zero once a token reaches sufficient uncertainty; an asymmetric augmented Lagrangian that permanently ratchets retain protection after any violation; and a bilevel structure confined to LoRA adapters that repairs retain damage before each forgetting step. BLADE dominates across three benchmark families, improving average composite scores over the strongest baselines by $6$% on TOFU, $9$% on MUSE Books, and $7$% on KnowUndo, and it remains stable under $4\times$ scaling and $4$ sequential unlearning steps on MUSE News where the best competing method collapses entirely.

The significant upsurge in vehicle traffic presents a considerable challenge in the pursuit of smart mobilization and transportation (SMT) worldwide. Current approaches primarily focus on vehicular traffic management through congestion prediction but fall short in addressing essential objectives such as traffic reduction and appropriate vehicle selection to alleviate congestion in smart cities ($SmCt$). To address these concerns, this work introduces a novel Neighbor-Embedded Graph Neural Network-based Crowd Delivery Traffic Management (NeCDM) Model, comprising two key components: the Traffic Congestion Prediction Unit (TCPu) and the Traffic Observation and Management Unit (TOMu). The TCPu utilizes Graph Neural Network (GNN) optimization to accurately predict traffic flow levels at various delivery stations within $SmCt$ ecosystems. Additionally, the TOMu facilitates the intelligent selection of the most suitable delivery vehicles for fulfilling crowd delivery requests ($CDR$). This work emphasizes the potential of crowd delivery as a feasible solution for achieving SMT goals while adhering to smart city parameters ($\mathcal{SCP}$s), such as reduced carbon emissions, shorter travel times, and minimized travel distances. The proposed model achieves notable improvements in computational efficiency, including reductions of up to 4.03% in L1 loss ($\pounds$), 16.66% in L2 loss ($\pounds_{rmse}$), and 7.64% in computation time.

Theoretical guarantees for kernel regression are typically formulated in terms of smoothness, but practical accuracy depends critically on how the design resolution compares with the target and kernel lengthscales. We develop a finite-sample theory for Matérn regression on periodic domains with quasi-uniform designs, covering noiseless interpolation and noisy kernel ridge regression. Our minimax result shows that accurate recovery requires a design dense enough to resolve the target lengthscale and sufficient information at that scale to overcome noise. We prove that Matérn interpolation additionally requires the design to resolve the kernel lengthscale: if the kernel is too short relative to the point spacing, an additional error remains even when the target is well resolved. For noisy Matérn regression, we derive a three-term fixed-ridge risk characterization consisting of target-scale bias, kernel-scale bias, and variance, and show that optimizing over the ridge parameter yields four distinct contributions. When the target is no more than twice as smooth as the kernel, choosing a kernel lengthscale longer than the target does not worsen the oracle risk, whereas choosing one too short can. Thus these resolution conditions determine when accurate recovery becomes possible, while smoothness determines how rapidly the error decreases thereafter.

Decentralized federated learning (DFL) is a promising paradigm for autonomous nodes to collaboratively train AI models without relying on a central server. However, existing DFL solutions do not guarantee global model consistency, a critical requirement for collaborative mission-critical scenarios where model divergence undermines decision uniformity and safety. This lack of consistency also amplifies vulnerability to Byzantine adversaries, who exploit the decentralized network topology and weak synchrony to perform equivocation and model poisoning attacks against individual victims. This paper introduces DFL-C, a novel Byzantine-resilient DFL architecture that enables decentralized nodes to perform collaborative training with global model consistency. At its core, DFL-C integrates an asynchronous common subset (ACS) consensus protocol into the DFL workflow to ensure all nodes aggregate a uniform set of model updates to establish global model consistency, despite individual Byzantine equivocation. DFL-C further implements a dual-domain trust scoring mechanism to provide resilience against data-domain Byzantine manipulations including model poisoning attacks. This mechanism complements the consensus protocol, significantly reducing the latter's runtime. Our experimental results demonstrate that DFL-C maintains model accuracy while achieving global model consistency under Byzantine behaviors with moderate consensus overhead. Notably, when compared with the state-of-the-art DFL solution BALANCE (Fang et al.) that does not provide model consistency, DFL-C achieves better model accuracy against untargeted model poisoning attacks and comparable resilience against backdoor attacks, with the advantage widened under non-IID scenarios.

Modern analytics is increasingly organized as recurring SQL pipelines rather than isolated SQL statements. Tools such as dbt, which have gained extreme popularity in recent years, allow teams to write each transformation as SQL and make dependencies between transformations explicit, producing directed acyclic graphs (DAGs) with hundreds or thousands of interdependent SQL models. Traditional query optimizers and source-to-source query rewriters operate on one query at a time, while materialized-view selection and multi-query optimization address narrower forms of reuse. They do not exploit the pipeline-level information exposed by explicit dependencies: how intermediate results are consumed, which downstream outputs depend on each computation, where expensive work sits relative to data reduction, which results are worth persisting, and how refresh schedules relate to input change and output demand. We introduce DAGSmith, to the best of our knowledge the first holistic dependency-aware source-to-source rewriting system for SQL pipeline DAGs. DAGSmith treats explicit dependencies as optimization signals. It analyzes each transformation with its upstream inputs, downstream consumers, and position in the pipeline DAG, uses an LLM to propose pipeline-level refactorings, separates SQL generation and equivalence checking to reject unsafe rewrites, retunes persistence choices with a learned cost model, and selects a globally compatible, conflict-free set of rewrites. This enables dependency-edge simplification, non-local semantic reuse, downstream-aware pruning, pipeline-aware work placement, rewrite-materialization co-optimization, and frequency-aware optimization. On the open-source Tuva dbt project, DAGSmith reduces elapsed time by 42.6% and warehouse compute cost by 67.7%, 98.1%/348.3% larger than state-of-the-art single-query rewriting.