Papers

Reporting the Fisher geometry of a trained variational quantum model is routine; quoting the shot budget that would establish it is not. Certifying an empirical Fisher matrix to relative Frobenius error $\varepsilon$ under coordinate-wise parameter shift costs $\Theta(B p^{2} V/(\varepsilon^{2} G))$ circuit executions, where $V$ is the measured readout variance and $G$ the measured squared gradient norm, with uniform allocation optimal in that class. One constant reproduces the cost of two circuit families whose exponents differ by a full power of $p$. The exponent is an identity in how $nV$ and $nG$ scale with the register, holding family by family to $0.001$ across 624 matrix-product-state cells once the finite-$p$ prefactor is removed. The cubic cost is therefore a finite-size window, set by whether the readout light cone grows with the register. A product family to 256 qubits gives $1.966$ (95% CI $1.934$–$1.997$); a brickwork entangler falls from $2.853$ below ten qubits to $1.715$ beyond sixty-four; a blocked entangler gives $1.984$ at a fixed cone width against $3.034$ at a proportional one. Fixed device connectivity fixes the cone, so a cubic budget from a small simulation overestimates a large machine, on top of hardware multipliers $2.07\times$ ($1.41$–$3.02$), $2.38\times$ and $1.91\times$ on ibm_marrakesh, ibm_fez and ibm_kingston. Cost-optimal readout weights cut the measured shot budget by $2.67\times$ ($1.33$–$4.00$) on hardware, flat from four to twelve qubits. A discrepancy model fitted on cheap circuits transfers its mean inside the calibration grid and, at six larger sizes named before the data, does not: nominal 90% intervals cover 36%, and split conformal is the only rung that stays near nominal.

Recent advances in multimodal foundation models have intensified the need to understand how different modalities share, preserve, and complement information. Mutual Information (MI), the Information Bottleneck (IB), and Partial Information Decomposition (PID) provide complementary perspectives, yet existing studies often treat them as isolated tools. This survey presents an information-theoretic perspective connecting these principles as progressively refined views of multimodal information processing: MI characterizes inter-modal dependency, IB explains task-oriented information preservation under compression, and PID decomposes preserved information into redundancy, uniqueness, and synergy. We review 170 recent studies (2018–2026) and 12 foundational works, organizing multimodal learning around four challenges: cross-modal alignment, information-efficient fusion, interaction-type characterization, and scaling to multimodal foundation models. Rather than using application domains as primary taxonomy axes, we interpret healthcare, robotics, recommendation systems, affective computing, and wireless communications as empirical validations of these information principles. Beyond taxonomy, we organize existing multimodal paradigms within a single information-theoretic coordinate system – the Generalized Multimodal Information Lagrangian – in which they occupy exact or approximate parameter corners, and whose unoccupied regions name candidate method families the literature has not yet built. We further discuss how emerging multimodal foundation models instantiate these principles at scale and identify open challenges including scalable information estimation in high-dimensional settings, standardized evaluation across information-theoretic methods, combinatorial complexity of multimodal PID, and the transition from post-hoc information analysis toward information-aware multimodal learning.

Many engineering building blocks behave as multi-port linear time-invariant systems. RF cavities, photonic devices, and superconducting quantum chips, despite their different underlying physics, all share a common mathematical structure for their port-level response. Each entry of the response matrix is a sum of contributions from a small number of intrinsic resonant modes, the pole-residue form. A model capable of predicting such responses for arbitrary geometries and arbitrary port configurations, while simultaneously extracting the underlying eigenmode structure, would therefore establish a foundational design principle spanning all these domains. We propose a neural framework that learns this modal decomposition end-to-end, supervised only by system-level observables and without supervising the modal parameters themselves. The architecture decomposes into a port-independent pole predictor and two port-dependent coupling predictors whose outputs are combined entry-wise, separating intrinsic from port-dependent features. This factorization yields a single trained model that generalizes to port counts unseen during training, dissolving the $\mathcal{O}(N^2)$ scaling barrier of direct regression. Despite no modal supervision, the freely-parameterized poles converge to physically meaningful eigenmodes, verified by cross-validation against the AAA rational approximation algorithm. We instantiate the framework in radio-frequency electromagnetic surrogate modeling. A model trained only on 2-port data accurately predicts $N$-port responses unseen during training.

Reliable small-molecule identification often requires complementary evidence from multiple spectroscopic measurements. In practice, however, spectra may be unavailable, degraded by measurement-related variations, or even incorrectly associated with a sample, thereby hindering accurate molecular identification. Herein, we propose a multimodal mixed-condition training strategy that accommodates missing, degraded, and mismatched measurements for small-molecule structure identification. The strategy incorporates chemical and spectroscopic knowledge through predefined missing-input configurations, modality-specific spectral perturbations, and chemically informed spectrum replacements. Models were trained on 635,441 samples comprising mass spectrometry (MS), infrared (IR), and nuclear magnetic resonance (NMR) simulated spectra from the Multimodal Spectroscopic Dataset (MSSD). They were then systematically evaluated on 79,462 held-out samples across 30 views designed to represent variations in spectra. A controlled comparison of complete-input and mixed-condition training under concatenation and mixture-of-experts (MoE) fusion showed that the training strategy was the principal source of improvement. For MoE, mixed-condition training increased the mean reciprocal rank (MRR) by 6.08% (from 0.9203 to 0.9763) and the top-1 molecular identification rate by 7.67% (from 89.50% to 96.36%). Notably, under single-modality inputs, IR MRR increased 2.15-fold (from 0.4337 to 0.9307), while MS MRR increased 2.31-fold (from 0.3711 to 0.8575). With the proposed strategy, complete-input performance remained high, while sample-level mismatch detection also improved. Together, these results highlight the potential of multimodal mixed-condition training for practical molecular identification by explicitly addressing incomplete, degraded, and mismatched measurements encountered in real-world analysis.

Decentralized training enables large-model training over low-end GPUs and internet-grade connections, but communication along both data-parallel and pipeline-parallel axes becomes the primary bottleneck. We study post-pretraining adaptation in this setting. We propose an asynchronous two-circuit system: a fast compressed training circuit drives throughput using activation masking for pipeline-parallel (PP) transfer and compressed data-parallel (DP) synchronization, while a slow anchor circuit runs occasional unmasked forward–backward passes off the critical path. Then, we introduce a spectral correction optimizer that uses these delayed anchor priors to denoise masked gradients without blocking the fast stream. Although prior work has found aggressive activation compression unreliable, we show that masking supports post-pretraining adaptation at high compression rates when anchored this way. Pipeline-parallel compression alone yields up to a $9\times$ throughput gain, and combining it with data-parallel compression increases beyond $40\times$ over internet-grade $\sim 200$Mbps connections, while matching dense uncompressed performance across domain adaptation and continual pretraining.

Variance penalization is a principled approach to risk-sensitive reinforcement learning (RL) that explicitly trades expected return for policy stability. Existing methods require a dedicated second critic to estimate return variance online, adding architectural complexity and compounding estimation error during learning. We propose a nonparametric variance-penalized actor-critic (VPAC) framework that replaces the variance critic with statistically grounded online estimators based on bootstrapping and random scaling, techniques drawn from the statistical inference literature for stochastic approximation. These estimators require no auxiliary network, maintain a single-critic architecture, and produce variance penalties that are bounded by construction, enabling clean convergence analysis. We establish almost-sure convergence for both a variance-penalized Q-learning algorithm and a two-timescale actor-critic variant via the ordinary differential equation (ODE) method, requiring only that variance estimates remain bounded rather than consistent. Empirically, we evaluate across discrete and continuous stochastic environments, demonstrating that the proposed methods match or exceed the variance reduction achieved by the existing dual-critic VPAC baseline while eliminating the overhead of a second critic. We further validate on a high-temperature superconductor (HTS) manufacturing case study, where VPAC-RS (Random Scaling) achieves a 74% reduction in steady-state critical current variability and a 63% reduction in episode return standard deviation, translating directly to improved yield consistency. Our results establish nonparametric statistical inference as a practical and theoretically sound alternative to auxiliary critics for risk-sensitive RL.

Against the backdrop of digital transformation and stricter regulation, enterprise compliance work demands higher efficiency and accuracy. The auxiliary compliance process has become an important entry point for optimizing the compliance system due to its strong transactional nature and high degree of repetition. This study focuses on the process characteristics of auxiliary compliance work, sorts out its structural composition and organizational mechanism, proposes an optimization path with process reengineering, system modeling, and technology integration as the core, and focuses on exploring the collaborative application of key technologies such as RPA, rule engine, and semantic recognition in process automation. Research suggests that the systematic optimization and intelligent upgrading of auxiliary processes will help build a modern compliance operation system that is responsive, efficient, structurally clear, and risk controllable.

Robust point tracking in endoscopic videos is essential for computer-assisted intervention and autonomous robotic surgery, enabling continuous registration between intraoperative video and preoperative imaging despite soft tissue deformation. However, supervised tracking methods depend on large annotated datasets, while surgical conditions make reliable trajectory annotation challenging. We propose a self-supervised Track-Any-Point approach that learns from unlabeled surgical videos by establishing global pixel correspondences and inferring point trajectories through contrastive random walks. Trained without annotations, our method achieves performance comparable to existing semi-supervised approaches while implicitly handling tissue deformation. These findings demonstrate the feasibility of self-supervised point tracking in surgical environments and its potential to reduce reliance on annotated data.

Predicting under an acquisition budget requires choosing feature groups whose value can depend on later queries. O-MPAC transfers finite-horizon risk-cost targets from complete training records into a shared source-action scorer. At inference time, the scorer uses partial observations and source metadata, re-scores after each query, and applies a hard cost mask. We analyze how tied teacher targets and the remaining planning horizon affect the learned decisions. Uniform supervision over tied minima preserves the target distribution under source relabeling. In a five-seed routing experiment, it achieves 0.965 accuracy under both original and context-last orders. On six real tasks, validation selects H1 without action cross-entropy in all thirty splits. O-MPAC has the highest mean budget-integrated accuracy on five tasks against source-adapted GDFS, DIME, AACO+NN and a static policy.

Machine learning models are increasingly deployed in healthcare imaging pipelines for diagnostic support, and training-time attacks against them are a named sector-level concern: healthcare-sector guidance identifies model poisoning and adversarial attacks as threats requiring dedicated defenses, while federal policy directs expanded AI vulnerability-detection tooling to critical infrastructure operators such as rural hospitals. Spectral signature analysis and activation clustering are two established backdoor detection methods routinely evaluated as independent baselines, but their outputs are not ordinarily combined, and reported detection performance on medical imaging benchmarks remains sparse relative to the natural-image setting. This paper contributes three things: a score-level fusion rule combining per-class spectral ranking with activation-clustering flags into a single per-sample poisoning score and a model-level agreement statistic; an open-source implementation of the resulting eight-stage pipeline; and an evaluation of that pipeline against synthetically poisoned variants of a public medical imaging benchmark and CIFAR-10 at four poisoning rates (0%, 1%, 5%, 10%) over five seeds each, measuring each detector alone against the fusion. On the medical benchmark, the fused detector reaches AUROC >= 0.99 at every nonzero poisoning rate tested. On CIFAR-10, fusion does not uniformly help: at 10% poisoning, activation clustering's true-positive rate collapses to 0.000 and spectral AUROC independently degrades to near-chance (0.545), despite a 97.2% attack success rate confirming the backdoor was fully installed. The fused score, a weighted combination of both signals, inherits this joint failure. Detection output is expressed in NIST AI RMF Measure-function and MITRE ATLAS terms, so findings are reported in the vocabulary security and compliance teams already use.

Temporal knowledge graph embedding (TKGE) models infer missing facts in knowledge graphs that evolve over time. Many existing models use a single geometric space, which can limit their ability to represent diverse relational patterns, or treat entity representations as static. We propose Biquaternionic Space with Complex-valued Attention (BSCA), a TKGE model that combines circular and hyperbolic rotations within a unified biquaternionic framework. A complex-valued attention mechanism adaptively fuses time-conditioned and relation-conditioned entity representations, allowing them to vary with temporal and relational context. Experiments on five benchmark datasets show competitive performance across datasets, with the largest improvement on GDELT: BSCA achieves an MRR of 52.1%, compared with 38.1% for the strongest baseline in our comparison.

Ringdown gravitational waves from binary black hole mergers can be modeled as superpositions of quasinormal modes (QNMs), whose frequencies and excitation factors encode properties of the remnant Kerr black hole. Reliable extraction of multiple QNM components is challenging because of mode overlap and noise. We develop an autoencoder-based framework for multi-component QNM analysis, in which the latent space is trained to represent the physical parameters of individual modes, enabling waveform denoising and parameter estimation within a common framework. Using controlled model waveforms constructed as finite sums of Kerr QNMs with recently established high-precision frequencies and excitation factors, including their nontrivial spin dependence near resonant excitation, we assess the method across partitioned spin intervals. The model achieves good in-domain waveform reconstruction and parameter recovery for the two longest-lived components of eight-component input waveforms, while its performance degrades when the validation spins lie far outside the training range. In a selected spin interval, the framework also recovers the 32 parameters of an eight-component waveform with good overall agreement. These results demonstrate the feasibility of physics-informed autoencoder-based inference for a prescribed multi-component ringdown waveform family and motivate further tests with progressively more realistic signals.

Joint source localization and geoacoustic inversion requires optimizing an objective built from an expensive normal mode propagation model. Bayesian optimization (BO) with a Gaussian process (GP) surrogate can obtain accurate parameter estimates within a limited number of forward model evaluations, but its performance depends on the choice of kernel family. With few observations in a seven-dimensional search space, no single kernel can be expected to perform consistently well across individual inversions. To reduce this dependence, we use a weighted ensemble of GPs with different kernel families, allowing the surrogate to adapt to the observed objective without committing to one kernel in advance. The ensemble is combined with an optimum-conditioned acquisition function that determines where the expensive objective should be evaluated next. Experiments on simulated and measured SWellEx-96 data show that the resulting method achieves the lowest mean final objective among the considered BO strategies and reduces parameter estimation error on most coordinates. Ablation results further show that the ensemble provides robustness to kernel choice, while the acquisition function accounts for most of the optimization gain.

Recent robot learning paradigms increasingly rely on large offline datasets of robotic interactions to train control policies. Expressive generative models enable rich and multimodal action representations, expanding the capability of this paradigm for complex robotic control. However, policy improvement with multi-step generative actors remains challenging. In offline reinforcement learning (RL), incorporating value-based objectives along generative trajectories often introduces substantial training complexity, including backpropagation through time (BPTT), auxiliary architectures, or distillation losses. We propose Value-Guided Flow Matching (VGFM), a scalable offline RL framework that enables dense value-guided shaping within a flow-based policy while avoiding BPTT and additional algorithmic overhead. VGFM parameterizes the policy as a conditional flow-matching model in action (x-prediction) space, ensuring that each intermediate flow step produces a valid robot action that can be directly evaluated by a standard offline RL critic. This design allows value guidance to be applied at randomly sampled flow times without differentiating through the entire generative trajectory, while preserving inference-time flexibility by varying the discretization of the underlying flow ODE without retraining. Evaluated on robotic locomotion and manipulation tasks in OGBench, VGFM achieves strong performance across a wide range of tasks under rigorous evaluation protocols. With minimal hyperparameter tuning, these results demonstrate that VGFM provides a simple, scalable, and effective approach for expressive policy learning in long-horizon, goal-oriented robotic control.

Database management systems increasingly serve dynamic and exploratory workloads, yet many of their decisions still rely on low-level signals such as recency, frequency, and address locality. These signals capture how data was accessed, but not what is being examined or how an analytical focus evolves. We argue for treating workload semantics as a first-class control signal for anticipatory decision making. Central to this view, we introduce semantic locality and semantic trajectories, which capture relationships among nearby queries and how those relationships evolve across a session. We propose a framework that represents semantic context at the data, query, and session levels, models its evolution over time, and translates it into task-specific utility estimates. We instantiate this framework in semantic prefetching and semantic cache eviction, which share a semantic layer to make two separate decisions. Prefetching uses semantic trajectories to anticipate future accesses beyond what address-based locality can capture, while eviction uses semantic relevance to inform block replacement. These systems provide initial evidence that shared semantic context can support multiple DBMS components. We further outline how this principle can extend to other decisions and data systems, and discuss key challenges in representation, cost, adaptation, and evaluation.

Bird-strike mitigation at wind farms requires identifying distant birds that span only tens of pixels, so the classifier's input resolution N is a design variable, not a fixed specification. We study it with a factorial design over 14 side lengths N (16 to 224), six architectures, two training and evaluation regimes and 30 random seeds – 2,520 checkpoints and 5,040 evaluations – plus latency measured on an NVIDIA Jetson Orin Nano. Four results. (1) The selected N depends on the target: 0.90 is met on validation by ResNet50 at N=112 in an estimated 1.85 ms (0.8980 on test) and 0.95 by DINOv2-L at N=144 in 12.70 ms; changing the model buys more accuracy than raising N (+5.93 versus +2.33 points at N=112). (2) The benefit of lowering N depends on the assumed preprocessing path: N=224 -> 80 saves 13.5% when each individual is decoded from its own file but 46.2% when the detector decodes the 4K frame once; the Pareto set grows from 20 to 23 configurations. (3) Accuracy must be measured on the deployed engine: half precision costs ViT-S/16 alone 4 to 7 points at N>=96 while the CNNs stay within 0.1 points, and with selection held at validation the choice differs at 26 of 176 targets. A broken FP16 engine can run faster than a correct one, undetectable from latency; admitting 14 ViT-S/16 FP32 configurations moves the recommendation over the 0.931-0.938 band and under the 10 ms budget. (4) ViT-L-scale models fit this device, but activations exceed the FP16 range; splitting the graph at transformer-block boundaries confines FP32 to the affected segments, making the deployed DINOv2-L chain 1.85x faster than the single-engine build. We also quantify how the regime-difference sign stabilises with seed count; a sensitivity split removing some forms of group sharing preserves all 14 non-trivial signs at the selection boundary.

In wireless federated learning (FL), data heterogeneity and multiple local updates induce client drift, degrading model convergence. It is further affected by unreliable wireless links, as transmission errors may invalidate model updates. To address these challenges, we propose a federated random walk averaging (FedRW) framework, which is a variant of federated averaging (FedAvg) that mitigates data heterogeneity by updating models along random walk (RW) paths and aggregating them at the server. Model parameters are transmitted in packets with retransmission support to improve training quality by mitigating wireless errors along RW paths. Meanwhile, wireless transmission delays hinder the exploration of FedRW. To this end, we formulate a joint optimization problem that integrates learning, RW path selection, and transmission parameter tuning, aiming to minimize the training loss under delay constraints. By deriving an upper bound on the expected convergence of FedRW over unreliable wireless networks, we reduce the problem to a general form agnostic to task type and model architecture. A distributed solution is then proposed, in which the server or clients optimize packet size and maximum number of retransmissions locally, and efficiently select reliable and expandable next-hop nodes via a resilience-aware beam search with dynamic pruning. Simulation results show that FedRW achieves 2.26%-9% higher accuracy than state-of-the-art baselines under high data heterogeneity. Furthermore, the jointly optimized FedRW yields at least 2.78% higher accuracy and faster convergence compared to baselines.

Static benchmarks for computer-use agents fix a task set at release and score every system against it once. That makes them reproducible, and it lets them drift from what they should measure: a fixed task set ages, leaks into training corpora, and cannot follow how people actually use agents from week to week. CoArena measures use directly. Real users submit tasks; two systems, each a single model or a multi-agent pipeline behind the same tool interface, execute the same task concurrently in identical sandboxed desktops; users judge the two outcomes without knowing which system produced them; and a public leaderboard is refit from those judgments. The central contribution is a formal account of what makes such an evaluation real-time. We define real-time as five measurable properties, each with an equation and a worked example: continuous task arrival, live concurrent execution, online rating updates, freshness with contamination resistance, and bounded feedback latency from a failed run to a reusable training environment. The rating methodology follows in full: the Bradley-Terry pairwise model, its likelihood with weighted observations and ties, the penalized maximum-likelihood estimator, and the streaming update applied when a single vote arrives (a stochastic-gradient step on the same likelihood, recovering Elo). It gives confidence intervals from the observed information and a cluster-robust sandwich, rank bands from a parametric bootstrap, the rule by which a new system enters the board, and the convergence rate of the estimate. Vote quality is treated with inter-judge agreement statistics, redundant judging, and explicit handling of ties and abstentions. A five-system example with 211 votes is carried from the vote matrix to ratings, intervals, and rank bands. Every number is derived from stated inputs or labeled illustrative; none is a measurement of a deployed system.

Financial fraud in corporate transaction networks has grown more coordinated and harder to detect with rule-based engines and with classical learning models that treat each transaction in isolation. This paper presents GTFD, a graph-transformer fraud detector that fuses structural and temporal evidence from a corporation's payment graph. GTFD encodes the graph with a multi-head graph attention network, encodes ordered transaction sequences with a gated transformer, and combines both views through a cross-modal gating layer. A conformal risk-control head converts the fused representation into threshold-free anomaly scores with finite-sample coverage guarantees, and the network is trained with self-supervised link-mask pretraining plus adversarial augmentation so it remains stable under scarce labels and under adversarial perturbation. On a corporate transaction benchmark enriched with coordinated fraud rings, GTFD reaches an AUROC of 0.990, an F1-score of 96.1% (precision 96.3%, recall 95.9%), and an accuracy of 98.4%. It reduces the false-positive rate by about 29% relative to the strongest baseline while raising coordinated fraud-ring recall from 85.1% to 96.5%. Ablations attribute roughly 2.0 AUROC points to self-supervised pretraining and 1.9 AUROC points to the conformal head, and adversarial stress tests show GTFD retains 89.2% accuracy at perturbation magnitude 0.20 where the next-best model falls to 76.4%.

Modern mobile inference runs on heterogeneous platforms combining mobile GPUs with multiple CPU core clusters. Existing optimizations typically exploit either inter-operator parallelism, by assigning entire operators to CPU cores or to the GPU, or intra-operator parallelism, by partitioning each operator for CPU-GPU co-execution. We consider these two forms of parallelism together, to improve inference latency of tasks that can be represented by a static DAG of operators with predefined input/output tensor shapes (e.g., CNNs or vision transformers). We define the problem of partition-aware DAG scheduling for mobile heterogeneous inference, illustrating that the best strategy depends on the structure of the inference DAG, thus motivating a joint formulation capturing operator partition choices, device assignment, and execution order. We propose an online iterative search framework, which decomposes large DAGs into stages, focuses search on critical operators, and uses latency predictors to estimate partitioned execution without exhaustive profiling. Across representative mobile inference workloads, our approach achieves latency close to an offline solution while keeping scheduling overhead to a fraction of the model initialization cost, allowing platform-specific scheduling at deployment time.

Open-source video-game quality depends on inter-actions among code, assets, configuration, tests, contributors, and issue workflows, yet conventional defect predictors usually flatten or omit these relations. We investigate release-level forecasting of a quality incident within thirty days using GAMEQUALGRAPH-Pilot, a typed temporal feature pipeline with calibrated risk estimates and effort-aware ranking. Because the accessible OS-SGameBench materials do not provide manually audited release dates and outbreak labels, the executed evaluation is explicitly simulation-backed rather than an empirical claim about real games. Five seeded worlds each contain 120 projects and 24 releases, with project-disjoint validation and future cross-project testing. The pilot obtains an AUPRC of 0.520, AUROC of 0.673, Brier score of 0.207, and 29.68% effort-aware recall at a twenty-percent testing budget. Its closest local comparator, Static-Hetero-Reimpl, reaches 0.522 AUPRC; the -0.002 difference is not statistically significant after Holm correction. Inference requires 0.023 milliseconds per release in the measured environment. Ablations and controlled missingness, drift, engine, project-size, alert-threshold, and attribution analyses expose where typed interactions help and where they fail. Results support the reproducibility of the proposed protocol, not deployment effectiveness. Real OSSGameBench release reconstruction, stratified label audits, and official graph-model comparisons remain mandatory before journal submission or operational use in practice. This boundary protects research integrity and supports credible evaluation.

Low-cost indoor air-quality sensors could support personalized asthma prevention, but their nonlinear measurement error, delayed exposure effects, time-varying confounding, and heterogeneous lower-tail responses limit risk estimation. We present CAUSALQUANT-ASTHMA, a measurement-error-aware causal quantile distributed-lag framework for short-horizon peak expiratory flow analysis. Sparse reference measurements train a nonlinear calibration model; stabilized sequential generalized-propensity weights address measured exposure assignment; and a susceptibility-modulated, smooth, noncrossing quantile model estimates lag-specific and sustained-exposure contrasts. Because no authorized cohort simultaneously provided dense indoor sensing, reference co-location, and outcome-compatible longitudinal data, evaluation used five semi-synthetic panels with known counterfactual truth, 150 patients and 12,600 patient-days per realization. Across eight methods, CAUSALQUANT achieved a dose-response integrated absolute error of 0.304 plus or minus 0.094, improving 24.2 percent over the strongest measurement-error and propensity-weighted baseline. It also obtained the lowest overall pinball loss, 1.065, while maintaining zero quantile crossings and 78.1 percent coverage for the nominal 80 percent interval. Sensor calibration reduced held-out exposure RMSE by 33.7 percent. Stress tests quantified degradation under sensor noise, missing personal measurements, and hidden confounding. These findings establish methodological feasibility and reproducibility, not clinical effectiveness; prospective, governance-approved external validation is required before patient-level interpretation or deployment.

Automated incident management in large-scale microservice systems relies on learning robust representations from multimodal observability data, including metrics, logs, and traces. Although recent self-supervised frameworks enable unified modeling for anomaly detection (AD), failure triage (FT), and root cause localization (RCL), they often struggle with non-stationary temporal dynamics and heterogeneous service dependency structures. In this paper, we propose STAR, a Spatial-Temporal Adaptive Representation learning framework that explicitly addresses these challenges through adaptive normalizations. STAR introduces two tightly coupled mechanisms: Temporal Adaptive Normalization (TAN), which dynamically normalizes multivariate time series using multi-scale temporal context, and Spatial Adaptive Normalization (SAN), which performs structure-aware normalization over service dependency graphs. Unlike prior methods that treat normalization as static or task-agnostic, STAR formulates it as a learnable, context-conditioned transformation aligned with the intrinsic properties of microservice systems. The resulting adaptive representations are integrated into a unified self-supervised framework, enabling end-to-end unsupervised support for AD, FT, and RCL tasks. Extensive experiments on two real-world microservice benchmarks demonstrate that STAR consistently outperforms all state-of-the-art baselines, yielding significant and stable improvements across all three tasks. Our results highlight adaptive normalization as a principled and effective mechanism for robust multimodal representation learning in complex software systems.

Reinforcement learning (RL) is widely used in language-model post-training to maximize rewards assigned to individual model outputs, such as scores from binary verifiers or reward models trained on human feedback. However, applications such as synthetic-data generation, fairness-related constraint satisfaction, and policy exploration require controlling the distribution of outputs across model generations rather than only maximizing expected reward. We propose a general RL-based framework for Distribution Matching allowing matching the distribution of a latent categorical attribute of model outputs to a specified target distribution. Empirically, we demonstrate that dominant post-training recipes such as Group Relative Policy Optimization (GRPO) reduce output diversity by concentrating policy probability towards a single mode. Entropy regularization and sampling temperature can improve the spread of the distribution but have constrained effectiveness, limited to apply only in token space and toward uniform distributions. We show that prior work in this area is a specific case of Distribution Matching involving the $L_2$ divergence. We then propose reward functions for other divergences such as KL and Jensen-Shannon and motivate them with theoretical justification. Finally, we demonstrate the effectiveness of our approach on a set of experiments involving mathematical reasoning and programming.

A principal who deploys a binary classifier with an abstention option must decide which queries the mechanism abstains on. The right choice depends on the adversary. A gaming adversary already knows the classifier and tries to manipulate features across the boundary, so the principal does best by abstaining on queries close to that boundary. The same boundary-localizing rule is the worst possible choice against a learning adversary who does not know the classifier: each abstention now tells the adversary that the boundary is nearby, which is enough to drive a binary search. We analyze this tension. The two natural defenses, abstaining at a fixed rate and abstaining near the boundary, are Blackwell-incomparable: neither can be simulated by post-processing the other's responses. The number of queries needed to reconstruct the boundary to error $\eps$ is $\tilde\Theta(d/\eps)$ under the first defense and $\Theta(d \log(1/\eps))$ under the second, where $d$ is the VC dimension of the classifier family and $\tilde\Theta$ suppresses factors polylogarithmic in $d$ and $1/\eps$. The first rate is a worst case over query distributions; no reconstruction algorithm can close the gap at the distributions that attain it. We characterize the Pareto frontier between the two defense objectives, and confirm both rates on seven binary-classification tasks spanning tabular, image, and language-model-feature inputs: label-plus-counterfactual access extracts the boundary with up to $200\times$ fewer queries than a published label-only baseline.

In the improving multi-armed bandits problem, each of $k$ arms has an unknown nondecreasing, discretely concave reward curve $f_i$, and pulling arm $i$ for the $t$-th time yields $f_i(t)$. For sufficiently long horizons, Blum and Ravichandran (ALT 2025) proved that randomized algorithms achieve an $O(\sqrt k)$ approximation to the best single arm when the scale $m=f^*(T)$ of the optimal arm is known ($T\ge2k$), and $O(\sqrt k\log k)$ when it is not ($T>4k$), against an $\Omega(\sqrt k)$ lower bound. The logarithmic factor is unnecessary: a one-page probe-and-commit algorithm achieves competitive ratio $4\sqrt3\,\sqrt k$ for $T\ge2\lfloor\sqrt k\rfloor$, without any knowledge of the scale, and we determine the optimal ratio for every horizon, $\Theta(\sqrt k+k/T)$, also for unknown horizons. Without noise, no prior is needed at all: a random-marginal probing algorithm reading neither the scale $m$, nor the concavity-envelope exponent $\beta$ of Blum, Garicano, Ravichandran and Sharma (UAI 2026), nor the horizon $T$, achieves the optimal $\Theta(k^{\beta/(1+\beta)}+k/T)$ simultaneously for every $\beta$ and every horizon. Under the multiplicative noise model of Blum and Ravichandran, probe-and-commit keeps the same all-horizon order $\Theta(\sqrt k+k/T)$ without knowing the noise level (and $\Theta(\sqrt k)$ on the same range), but the price of priors jumps: for any fixed noise level $\varepsilon\in(0,1/2]$, the uniform price of adaptation $\phi_\varepsilon(k)$ — the worst case over horizons $T\ge16k$ of the loss relative to $k^{\beta/(1+\beta)}$ for algorithms knowing neither $m$ nor $\beta$ — is $\Theta_\varepsilon(\sqrt{\log k/\log\log k})$, the lower bound asymptotic in $k$ at fixed positive $\varepsilon$ and matched by a nested random-permutation probing algorithm, whereas knowing either $m$ or $\beta$ alone restores a constant price.

Can gradient-based training learn the rank needed to store and compose associations in a matrix memory? In our earlier study, we used a matrix-augmented reasoner on a task that admits a rank-1 solution, leaving this question open. We train matrix memories on $K$ fresh key-value bindings whose exact linear recovery requires $\mathrm{rank}(Z) \geq K$. A fixed linear readout queries a single matrix state without access to the original bindings. Experiments measure recovery by cosine similarity greater than 0.9, a threshold distinct from mathematical equality. Learned effective rank increases with $K$ across the tested grid (Spearman $\rho = 1.0$ at $d = 16$). Training-time rank caps produce a recovery transition near $k = K$: at $d = 8$, $K = 4$, rank 3 gives at most 0.0004 recovery and rank 4 gives 0.97. Four of five seeds retain at least 0.9996 recovery through 21-fold self-application of the trained operator. On the entity subspace, the learned operator has effective rank close to $K$ and approximates the ideal cycle. For the single converged seed capped below $K$, a calculation using the entity-subspace operator and ideal cycle predicts the measured cosine within 0.008 through seven applications. Extending training resolves several initial failures, but recovery still declines at larger matrix dimensions with encoder width fixed.

On-policy distillation (OPD) is increasingly applied to frontier foundation model post-training. Prior work in this area has largely focused on algorithmic advances, yet it remains unclear how much OPD depends on its training questions and, in particular, how far this dependence can be reduced. Across two representative single-teacher OPD settings, we find that training on 8 real prompts yields performance comparable to training on 17k problems, while datasets differing substantially in measured difficulty and initial distillation gap yield similar outcomes. Our analyses suggest two complementary explanations: repeated sampling could allow even a few prompts to expose substantial teacher supervision, while OPD transfers generalizable reasoning capabilities beyond dataset-specific knowledge. Building on these observations, we next propose a data-free on-policy distillation (DF-OPD) setting to investigate whether the system can supply the training questions itself, eliminating the need for external data. With 64 self-generated questions obtained without seed examples, DF-OPD yields performance comparable to full-data OPD in both single-teacher settings. This finding also holds in multi-teacher OPD: across mathematics, code, and instruction following, 1k generated questions achieve performance comparable to training on approximately 7k real post-training examples. We further explore whether OPD can operate even without explicit training questions. The experiments show that this is effective only in limited cases, where the student unexpectedly generates and answers its own questions, thereby reducing the process to an implicit form of DF-OPD. Together, these findings invite a reassessment of the role of training data in on-policy distillation. Code is available at https://github.com/Ryuki661/DF-OPD

Memory-efficient feature representations are increasingly important in machine learning settings where storage, transmission cost, bandwidth, or privacy constraints limit access to raw data. Bloom Filter (BF) encodings provide compact probabilistic representations of engineered features, but their behavior under structural compression and their applicability to regression tasks remain underexplored. In this work, we propose entropy-punctured Bloom Filters, a memory-aware encoding strategy that removes low-variability bit positions identified using empirical entropy. Starting from fixed-length BF encodings of quantized features, the proposed approach produces reduced representations that preserve predictive structure while improving predictive efficiency relative to encoded representation size. We evaluate the approach on diverse regression datasets, comparing raw features, Principal Component Analysis (PCA), Random Projection (RP), and Bloom Filter variants under leakage-free evaluation protocols and approximately matched representation sizes. Performance is assessed using ridge regression, XGBoost, and neural networks, with predictive efficiency measured as R2 relative to encoded representation size per sample. Results show that Bloom Filter encodings remain competitive with classical compressed representations while achieving substantial storage savings. Entropy-based puncturing further reduces representation size with minimal loss in predictive fidelity, yielding improved predictive efficiency. These findings demonstrate that entropy-punctured Bloom Filters provide an effective representation-level compression approach for memory-constrained machine learning.

Neural Architecture Search (NAS) automates network design, but evaluating a single candidate requires training it to convergence, making exhaustive search intractable. Zero-cost proxies estimate architecture quality at initialization in seconds, yet a single proxy is noisy, and combining several does not straightforwardly help: proxies are strongly correlated, so naive aggregation compounds their shared errors instead of averaging them out. Existing methods exploit either proxy signals or architectural topology - never both within a single active-learning framework. We introduce ZAPS (Zero-cost Active Proxy Search), a four-stage pipeline that closes this gap. ZAPS (i) selects a compact, non-redundant proxy subset offline via ProxyFit, a greedy anti-redundancy criterion; (ii) seeds the search with a hybrid K-means strategy that balances exploitation and exploration; (iii) re-selects proxies at every iteration by a bootstrapped vote as the labeled set grows; and (iv) ranks candidates with an XGBoost ensemble trained jointly on proxy ranks and one-hot topological encodings, queried through an Upper Confidence Bound (UCB) acquisition function. On NAS-Bench-201 under a budget of B=200 evaluations, ZAPS recovers 52.3% of the true top-100 architectures on CIFAR-10 and 65.8% on CIFAR-100, ahead of every baseline we consider - Random Search, Local Search, REA, BANANAS and TPE - and, on CIFAR-10, with less than half the run-to-run standard deviation of the strongest of them. The advantage is largest where evaluations are scarce: on NAS-Bench-201 it narrows as the budget grows, whereas on the harder NAS-Bench-101, which no method comes close to saturating, it widens instead. All methods are scored by a single criterion: how much of the true top-100 lies among the architectures they actually evaluated.