Papers

One-shot federated graph learning (FGL) requires the server to estimate client contributions from highly compressed information, yet conventional volume-based weighting captures the amount of client data while overlooking how its connectivity is organized. In this paper, we propose SPIRE, a Structural Entropy-Driven Graph Diffusion Generation method that introduces topology-aware client differentiation into one-shot FGL. Specifically, we employ first-order degree-distribution structural entropy as a compact descriptor of degree-mass dispersion and use it to derive structural client weights, providing an inductive bias that accounts for differences in graph topology beyond data volume. On the generation side, a graph diffusion model on the server synthesizes pseudographs conditioned on the weighted client prototypes, capturing both semantic and structural information without requiring additional client-side training. The generated pseudographs are then assembled via disjoint union fusion to train a global graph neural network. Extensive experiments on seven real-world graph datasets demonstrate that SPIRE consistently outperforms conventional and one-shot FGL methods, with particularly strong gains under highly heterogeneous (non-IID) and graph-perturbed settings.

Recursive Super-Resolution (SR) extends fixed-scale SR to extreme magnification by repeatedly feeding predictions back into the same model, analogous to zooming an image repeatedly. However, ground truth availability at every scale, especially at depth, remains challenging as the required source resolution grows geometrically, leaving deeper predictions unsupervised. We present OracleZoom, an on-policy distillation-inspired, reference-constrained framework that trains on its trajectory while carrying the last ground-truth evidence beyond the supervision boundary. Direct and cross-scale supervision constrain verifiable content, while a no-reference quality objective guides unresolved fine-scale detail. A KL-constrained pretrained latent prior limits quality-driven drift, while EMA consistency stabilizes the supervision boundary. Across seven datasets, OracleZoom achieves the state-of-the-art SR quality across zooming scales, averaging 0.713 CLIPIQA, with larger gains on deeper scales, while significantly reducing hallucinations. Code, data, and models are available at https://dipta007.github.io/OracleZoom/ .

Monte Carlo Tree Search (MCTS) has demonstrated success in online planning for deterministic environments, yet significant challenges remain in adapting it to stochastic Markov Decision Processes (MDPs), particularly in continuous state-action spaces. Existing methods, such as HOOT, which combines MCTS with the Hierarchical Optimistic Optimization (HOO) bandit strategy, address continuous spaces but rely on a logarithmic exploration bonus that lacks theoretical guarantees in non-stationary, stochastic settings. Recent advancements, such as POLY-HOOT, introduced a polynomial bonus term to achieve convergence in deterministic MDPs, though a similar theory for stochastic MDPs remains undeveloped. In this paper, we propose a novel MCTS algorithm, \Algname, designed for continuous, stochastic MDPs. \Algname integrates a power mean as a value backup operator, alongside a polynomial exploration bonus to address the non-stationarity inherent in continuous action spaces. Our theoretical analysis establishes that \Algname converges at a polynomial rate of $\mathcal{O}(n^{-\zeta})$, $\zeta \in (0,1/2)$, where \( n \) is the number of visited trajectories, thereby extending the non-asymptotic convergence guarantees of POLY-HOOT to stochastic environments. Experimental results on stochastic tasks validate our theoretical findings, demonstrating the effectiveness of \Algname in continuous, stochastic domains.

Planning with a generative model aims to estimate the value of a state using as few simulator calls as possible. SmoothCruiser achieves problem-independent complexity $\widetilde O(\varepsilon^{-4})$ by exploiting the smoothness of the entropy-regularized Bellman backup, but its estimator is only first-order. We show that the sample-complexity exponent of SmoothCruiser-type planners is governed by the order $\beta$ of the local Taylor remainder, giving oracle complexity $\widetilde O(\varepsilon^{-(2+2/(\beta-1))})$: the first-order case $\beta=2$ recovers SmoothCruiser, while a second-order/cubic remainder $\beta=3$ yields $\widetilde O(\varepsilon^{-3})$. We reach this regime with an optimal-transport-smoothed Bellman backup over action distributions, which has a closed form, a policy gradient, and a Lipschitz Hessian, and whose quadratic correction admits an unbiased cross-product estimator. The resulting SecondOrderSmoothCruiser achieves $\widetilde O(\varepsilon^{-3})$ oracle complexity for fixed OT parameters, and we relate the OT, entropy-regularized, and unregularized objectives through explicit regularization-bias bounds.

This study focuses on the scientific problem of understanding internal factors that govern the diverse performance degradation of deep neural networks (DNNs) when different parameters are pruned. In order to explain why pruning certain parameters leads to significant performance degradation but pruning other parameters does not, we examine how the pruning operation affects the interaction patterns encoded by the DNN. We find that when we progressively increase the pruning ratio, the interaction patterns encoded by DNNs exhibit a distinct three-phase dynamics, i.e., model performance is not largely affected until the pruning operation begins to remove low-order interactions, and low-order interactions exhibit strong generalizability. Moreover, we find that the high sensitivity of DNN performance to the pruning of certain modules is attributed to whether the pruning operation removes generalizable low-order interaction patterns.

Kernel methods separate data representation from decision-making, but typically require the kernel to be chosen in advance. We show that this kernel can instead be learned by alignment, and develop the resulting framework through the recently introduced Collaborative Learning and Inference (CLaI). We show that Collaborative Learning can be viewed as a kernel alignment process, in which an embedding is trained so that its induced similarity matches a label-derived target kernel. We also prove that Collaborative Inference is equivalent to kernel Bayes classification with Parzen-window density estimation. Motivated by these perspectives, we generalise CLaI by replacing cosine similarity with a learned Mahalanobis distance and extend it to multiclass classification. On CIFAR-10, PathMNIST, and SleepEDF, the Mahalanobis formulation improves accuracy, converges faster, and yields lower calibration error than the cosine-based variant. Auxiliary experiments further support these connections, showing that CLaI produces latent signals of the same form as a Gaussian process, while achieving competitive calibration on sepsis prediction. Together, these results establish a principled learned-kernel framework that unifies representation learning, kernel alignment, and Bayesian classification, and extends naturally to the multiclass setting.

Inference-time activation steering enables behavioral control of large language models without parameter modification, while post-training quantization reduces memory and compute costs for deployment. Despite their growing convergence in practice, the interaction between these two techniques remains uncharacterized. We systematically study activation steering under weight-only quantization (INT8 and NF4) across four open-weight 7-9B models and two behavioral targets: judged sentiment and judge-free reasoning length. Using an iso-effect framework that compares capability costs at matched behavioral effect, we find that sentiment steering survives quantization intact. After correcting a GSM8K parser artifact with a uniform v2.3.1 rescore, the pooled INT8 contrast is -0.010 (90% CI [-0.026, +0.007]), descriptively Equivalent under the preregistered three-label rule, while NF4 remains Inconclusive at -0.017 ([-0.067, +0.033]). In contrast, reasoning length exhibits a surprising asymmetric dose-response: lengthening is graded but terminates in cap-runaway and collapse, while shortening is a step function with only 12-30% shortening (model-dependent) before discontinuous failure. We expose a methodological pitfall: the naive iso-effect ladder anchors on the collapse floor for floor-bounded targets, and we introduce a censored construction that restores interpretable crossings. We also quantify a substantial baseline capability shift for Mistral-NF4 (0.545 to 0.365 GSM8K at alpha=0), demonstrating that compression can dominate the steering intervention. Despite this, steering vectors remain highly collinear with their FP16 siblings (cosine similarity 0.989-0.998 for INT8, 0.945-0.990 for NF4), confirming that the behavioral direction survives quantization even when the cost structure does not. All code and data are released.

Reinforcement learning (RL) across multiple domains can broaden the reasoning capabilities of large language models (LLMs), yet joint training often degrades individual-domain performance and can destabilize optimization. Existing work typically diagnoses such interference from a single-step view using first-order gradient alignment or curvature-based proxies. We show that this view can miss a critical form of sequential interference: same-point domain gradients may remain nearly orthogonal even when consecutive realized updates partially reverse one another in output space. We further show that consecutive token log-probability footprints recover this interaction directly from adjacent checkpoints as a local second-order interaction in output space, without explicitly reconstructing same-step curvature. Building on this insight, we propose OSOL, which designates a focus domain at each iteration, uses the preceding checkpoint footprint to rank token-level rebound risk, and applies a drift-ranked, adaptively scaled correction within the standard GRPO update. Our analysis shows that this correction suppresses the targeted cross-step output backtracking component. Controlled studies further show that cross-step backtracking is more strongly associated with subsequent task damage than same-point gradient diagnostics, while the preceding footprint ranks future rebound risk more accurately than Hessian-based proxies. On Qwen3-30B-A3B, OSOL reaches a domain-macro average of 0.4822, improving by 5.7% over the strongest compared baseline, without explicit higher-order differentiation.

K-Means is one of the most widely used clustering algorithms, but its susceptibility to initial centroid selection remains a primary bottleneck for its convergence speed and clustering accuracy. This paper proposes Sector-Mean Initialization, a deterministic initialization strategy with O(N) time complexity that partitions the two-dimensional data space into angular sectors around the global centroid and initializes centroids using sector-wise means. We evaluate the method on established two-dimensional benchmarks (SIPU, Birch) and multiple real-world datasets, comparing against random, K-Means++, and Max-Min initialization under identical Lloyd iterations. The statistical analysis of Friedman's test (p<0.05) and Nemenyi post-hoc comparison indicates that, while delivering equivalent clustering quality as K-Means++ and Max-Min, Sector-Mean offers significant computational efficiency. Experimental results show that Sector-Mean reduces the initialization time by 74.9% and 59.8% in comparison to K-Means++ and max-min, respectively. And, it yields the lowest average number of iterations, achieving approximately 5% fewer iterations than K-Means++ and 16% fewer than max-min. These results highlight that Sector-Mean initialization offers a deterministic and computationally efficient initialization strategy while preserving cluster quality.

In performative reinforcement learning the deployed policy shapes the environment that generates the learner's future data, and the natural solution concept is a performatively stable policy that is optimal in the environment it induces. Existing convergence guarantees rely on Lipschitz sensitivity assumptions on the environment map $\pi \mapsto (P_\pi, r_\pi)$, which are hard to verify and fail in settings such as multi-agent best-response dynamics. We instead study stability for mixtures of policies, and show that the resulting picture is fundamentally different from performative prediction, where randomization removes the need for any sensitivity assumption. We distinguish local mixed stability, an occupancy-weighted first-order relaxation that we show is equivalent to stationarity, from global mixed stability, which certifies against arbitrary deviating policies. Our first result is that a weighted per-state Hedge dynamic drives the local stability gap to zero at an $O(1/\sqrt{T})$ rate for an arbitrary, possibly discontinuous, environment map, both with exact and with trajectory feedback. The two notions genuinely differ: we exhibit an instance where local stability is achieved exactly but every mixture has global stability gap bounded away from zero. For global stability we introduce a bounded transition range assumption, strictly weaker than Lipschitz sensitivity, under which unweighted per-state Hedge converges up to a floor of $O(\gamma\epsilon_P/(1-\gamma)^3)$, and we prove a matching-in-$\epsilon_P$ lower bound of $\Omega(\gamma\epsilon_P/(1-\gamma))$ under trajectory feedback, so this floor is unavoidable. Finally, we extend both notions to $n$-player performative Markov games, obtaining local stability with no assumption on the joint environment map or game structure, and global stability for performative Markov potential games.

Recently, Masked Pretraining (MPT) based on reconstruction pretraining tasks has risen to a promising self-supervised learning paradigm across various domains and achieves remarkable performance in multiple downstream tasks. However, the theoretical understanding of the working mechanism behind MPT is still limited. In this paper, we introduce a new theoretical framework to analyze MPT and understand the crucial role of masking in extracting meaningful representations. We establish theoretical connections between MPT and another popular self-supervised paradigm: contrastive learning. We prove that the masking technique implicitly creates positive pairs that are semantically similar and the reconstruction loss pulls them together in the feature space. Besides, as a result of the implicit alignment, we point out the dimensional collapse issue of MPT and propose a Uniformity-enhanced MPT (U-MPT) loss that can effectively address this issue and bring significant improvements in downstream tasks including linear evaluation, cross-dataset fine-tuning and out-of-distribution generalization on real-world data sets. Furthermore, we establish downstream guarantees of U-MPT and theoretically analyze the influence of masking strategies. Based on the theoretical analysis, we propose a new masking strategy which enhances the downstream performance of MPT and explains current improvements of masking strategies with our theoretical perspective.

An LLM judge evaluates outputs at scale. Experts should label only where it is least sure. Its natural escalation signal conflates two uncertainties: aleatoric, real disagreement in the expert pool, which labels cannot reduce, and epistemic, the judge's ignorance, which labels do reduce. A small Bayesian model separates them: a regression on labels already collected learns how far to trust a black-box judge's prediction. Both components follow as simple formulas, with no sampling or further judge calls. The components isolate on a real LLM judge against exactly known truth, and stated confidence is no guide to its actual error. On real human disagreement (ChaosNLI) the epistemic ranking removes 83% more error than total uncertainty for the same expert labels, though simply escalating the least-labelled items does as well there. We demonstrate we can estimate where a judge is ignorant rather than where experts genuinely disagree, and propose using this to direct expert labelling. Code and data are available at https://github.com/composo-ai/judge-uncertainty-decomposition.

Diffusion-weighted imaging (DWI) is widely used in clinical settings but remains vulnerable to geometric distortion. Conventional correction methods often require additional acquisitions or vendor-specific solutions, limiting their feasibility in high-throughput, resource-constrained settings. This study investigates whether large-scale pretraining strategies can improve deep learning-based distortion correction for single-phase-encoding DWI. We formulate the task as image reconstruction, and compare a non-pretrained baseline against a self-supervised and a generative pretrained model, evaluated using both quantitative image-similarity metrics and qualitative expert assessment. The best-performing model was further tested for transferability on data collected in an LMIC setting with acquisition shift. Pretrained models outperformed the non-pretrained baseline, with cWDM achieving the strongest results across both quantitative and qualitative evaluation. However, application to LMIC data revealed transferability challenges, including contrast alteration and over-reliance on T1-weighted anatomical structure. Registering images to a common standard space improved predictions, suggesting that harmonized preprocessing may enhance cross-domain deployment.

We study the identity straight-through estimator (STE) for training a two-layer binary-activation network with hinge loss from the perspective of Statistical Learning Theory (SLT). Our central question is whether algorithmic stability can explain the statistical generalization of the estimator produced by the discontinuous STE training rule. In the saturated-output regime, the zero-initialized samplewise STE recursion is exactly the stochastic subgradient descent on the convex latent loss $(-yu^\top x)_+$. This representation makes a stability analysis possible. We derive an exact distance identity for two coupled updates and prove approximate non-expansiveness of the common-example map, with a quadratic defect only when the two latent margins straddle zero. We then obtain explicit $\ell_2$ on-average model-stability and generalization bounds, transferring stability isometrically from the latent vector to the full first-layer matrix. Combining stability with a standard optimization bound yields an explicit excess induced-risk guarantee and the rate $O(n^{-1/2})$ when $T=n^2$. Under margin separability, a complementary argument gives the optimal-order $O(R^2/(\gamma^2n))$ expected excess misclassification error for a randomized one-pass STE iterate and a corresponding majority-vote bound.

Small additive ensembles of symbolic rules offer interpretable prediction models. Traditionally, these ensembles use rule conditions based on conjunctions of simple threshold propositions $x \geq t$ on a single input variable $x$ and threshold $t$, resulting geometrically in axis-parallel polytopes as decision regions. While this form ensures a high degree of interpretability for individual rules and can be learned efficiently using the gradient boosting approach, it relies on having access to a curated set of expressive input features so that a small ensemble of axis-parallel regions can describe the target variable well. Absent such features, reaching sufficient accuracy requires increasing the number and complexity of individual rules, which diminishes the interpretability of the model. Here, we extend classical rule ensembles by introducing logical propositions with learnable sparse linear transformations of input variables, i.e., propositions of the form $\mathbf{x}^T\mathbf{w} \geq t$, where $\mathbf{w}$ is a learnable sparse weight vector, enabling decision regions as general polyhedrons with oblique faces. We propose a learning method using gradient boosting based on a weighted logistic regression. Empirical results across 14 regression and classification tasks demonstrate that the proposed method achieves lower model complexity than competitive baselines while maintaining similar or better predictive accuracy. Hence, the approach provides a favorable trade-off between interpretability and accuracy and reduces the reliance on manual feature engineering.

Batch normalization (BN) substantially improves sample efficiency in continuous-control actor-critic methods such as CrossQ, yet recent studies report performance degradation in discrete-action value learning on Atari. These failures are surprising because discrete Q-networks lack the action-input distribution mismatch identified by CrossQ. We show for target-based C51 and target-free PQN that the simple choice between running and batch statistics at specific forward passes can reverse this degradation. In C51, switching the BN bootstrap forward to batch-statistic mode significantly improves performance over unnormalized and LayerNorm baselines and scales stably with update-to-data ratios up to 12. In PQN, using batch-statistics for both action selection and bootstrapping recovers performance from the failing running-statistic configuration. Across 26 Atari games at 400M frames, this configuration achieves a higher final aggregate score than PQN with LayerNorm. Our results show that carefully configured BN can substantially improve discrete-action value learning, and that its forward protocols are an essential part of the algorithm specification.

Medical vision-language models (VLMs) require confidence that reflects both answer correctness and patient-specific visual evidence. Recent GRPO-based methods optimize verbalized confidence together with answer generation. However, this joint optimization may interfere with answer learning and drive confidence toward near-binary values. Verbalized confidence also provides no explicit assessment of visual support. We therefore separate capability learning from confidence estimation and propose DualRead. DualRead builds on the insight that reliability can be read from the actor's internal states at critical moments in the answering process. It freezes the GRPO-trained actor and combines pre-answer solvability with a post-answer assessment of the generated answer and its visual support. To further assess whether confidence reflects visual grounding, we introduce Counterfactual Confidence Grounding AUC (CCG-AUC). It measures whether confidence decreases when real-image substitution changes the actor from correct to incorrect. Across two VLM backbones and both in- and out-of-distribution medical VQA benchmarks, DualRead improves correctness discrimination and calibration over verbalized confidence while preserving answer accuracy. CCG-AUC reveals whether confidence responds to answer-relevant visual evidence rather than primarily to non-visual cues.

Recursive self-improvement (RSI) lets a system improve the model-building machinery from its own failures, so every later model inherits the gain. Yet RSI has been validated almost exclusively on coding and formal benchmarks such as science QA and mathematics. This format bound limits RSI to improvement within a machine-checkable slice, not general capability where questions are open and correctness is settled by argument, replication, or measurement. We argue RSI must next operate across real, diverse scientific, engineering, and meta-scientific domains, not where formal evaluation is merely tractable. To that end we present MetaRSI-v1, where improvement is the scheduled composition of three typed operators over one unified paradigm. Data-RSI amplifies existing competence and marks its boundary; Harness-RSI edits a five-slot scaffold without touching weights; Model-RSI internalizes capability into parameters through bounded training. Sharing one loop kernel and artifact vocabulary, they make data, scaffold, and model changes composable rather than exclusive. A two-axis optimizer jointly decides operator order and each operator's proposal policy, while a meta-level policy revises the schedule across terms. We validate MetaRSI-v1 under the field's standard evaluations, on code and closed-form science, with no external teacher: the target model plays every role in its own loop. MetaRSI-v1 reframes self-improvement from a single-surface edit to a composition across the full model-production pipeline, opening two paths: a model route internalizing capability through training, and a harness route leaving weights untouched and thus extending self-improvement to any model reachable through an interface, with Data-RSI redefined as the shared substrate feeding both. The framework further yields refutable laws on where loops exist, how operators compose, and what supervision buys.

Massive datasets in modern machine learning have made data reduction a central challenge, particularly for clustering tasks where memory and computational constraints demand compact yet faithful summaries. A standard approach is to construct an \textit{$\epsilon$-coreset}: a small weighted subset that approximately preserves the clustering cost for every plausible choice of centers. For the \textit{$(k,z)$-clustering problem}, existing worst-case bounds on coreset size are essentially tight, ruling out substantially smaller coresets in general. However, such worst-case instances are often unrepresentative of real-world data. In this work, we show that significantly smaller coresets are possible under mild and natural assumptions on the underlying data distribution. We introduce a new correlated sampling framework, called determinantal sampling, based on a novel application of determinantal point processes. Using this framework, we obtain an efficiently constructible $\varepsilon$-coreset for $(k,z)$-clustering in $\mathbb R^d$ whose dependence on $1/\varepsilon$ has exponent strictly smaller than $2$ when $d$ is fixed. This improves over the worst-case $\varepsilon^{-2}$ barrier under our beyond-worst-case assumptions. To the best of our knowledge, this is the first result that provably surpasses these lower bounds through beyond-worst-case assumptions. Finally, we validate our approach on synthetic and real-world benchmark datasets, where it consistently achieves smaller coresets than existing state-of-the-art methods, even without explicitly enforcing the assumptions used in the analysis.

Group-based reinforcement learning with verifiable rewards (RLVR) scoresmultiple completions per prompt using automatic verifiers. Analysesbased on independent verifier errors may overlook dependence associatedwith shared answer formats. We investigate this dependence in24,998 groups of eight completions generated by Qwen2.5-1.5B onMATH, GSM8K, and DeepMath-103K. We estimate a pooled within-groupverifier-error correlation of 0.530 (95% confidence interval:0.500–0.560). Under an exchangeable-error model, this correspondsto a design-effect-adjusted effective sample size of 1.70 for aneight-completion group. Dependence varies substantially across answerforms: fractions, radicals, symbolic expressions, and intervals exhibitstronger clustering than unit annotations and percent signs. Replayinggroup-relative advantages across four rule-based verifier configurationsidentifies at least one advantage-sign disagreement in up to 0.83% ofgroups. Because a group is repeated sampling for one prompt, thiswithin-group clustering may reflect shared prompt difficulty as well asshared answer form, and we do not attempt to separate the two here.Unlike studies of correlated judgments across multiple evaluators, ouranalysis examines dependence across completions scored by the sameverifier. These findings motivate prompt- and answer-form-aware analysesof verifier noise rather than characterizations based solely onaggregate error rates.

Motivated by the growing importance of fairness in machine learning, fair $k$-center clustering has attracted considerable research attention as a fundamental problem. In this problem, a dataset is partitioned into $m$ disjoint groups, and the objective is to select $k$ data points as centers, subject to upper bounds on the number of centers chosen from each group, aiming to minimize the maximum distance between any data point and its assigned center. Focusing on Euclidean spaces, which are ubiquitous in machine learning applications, we first develop a parameterized approximation algorithm for Euclidean fair $k$-center with an approximation ratio of $2.732$. By incorporating this algorithm as a post-processing stage into a one-pass streaming framework for large-scale data, we obtain an approximation ratio of $4.464$. These ratios can be further respectively improved to $2.414$ and $3.828$ with a runtime exponential on $k$. To ensure polynomial-time complexity, we further design a one-pass streaming algorithm with an approximation ratio of $4.732$, which can be further improved to $4.42$, outperforming the state-of-the-art ratio. Finally, extensive experiments show that our methods significantly outperform state-of-the-art approaches in terms of clustering accuracy.

In many scientific disciplines, weak signals of interest are obscured by dominant nuisance signals that are several orders of magnitude stronger. Recovering these weak signals requires subtracting the dominant ones; however, this calibration process inherently distorts or partially suppresses the underlying signal of interest. To address this problem, we propose the use of normalizing flow models to reconstruct calibration-affected weak signals. By leveraging the statistical invariance of the target signals and assuming minimal initial suppression, our framework effectively recovers the lost signal components. We provide a comprehensive theoretical overview of this normalizing flow-based recovery method and demonstrate its efficacy using simulated data.

LLM-as-a-Judge has emerged as a promising paradigm for evaluating natural language generation. However, the uncertainty associated with such evaluations remains largely unexplored, which limits their reliability in real-world applications. Although conformal prediction offers a principled framework for uncertainty quantification, existing approaches typically apply it to a single LLM judge, overlooking the variability introduced by using different LLM evaluators. In this work, we propose a robust uncertainty estimation framework for multi-agent LLM-as-a-Judge evaluation. Our approach constructs conformal prediction intervals for LLM-based scores from multiple LLMs. By considering intervals from different LLM judges, we obtain more stable and reliable uncertainty estimates. Extensive experiments demonstrate that our method produces valid prediction intervals with coverage guarantees, and that interval-based aggregation across multiple judges leads to more stable evaluation outcomes.

Hyperspectral image (HSI) classification plays a vital role in remote sensing applications, including agriculture, environmental monitoring, and urban analysis. However, its performance remains challenged by high spectral redundancy, noise sensitivity, and the difficulty of jointly modeling local material composition and long-range spectral dependencies. To address this, we propose AGSA-Net, an abundance-guided self-attention network that explicitly integrates spectral unmixing priors into the classification process. AGSA Net first estimates physically meaningful subpixel abundance maps subject to non-negativity and sum-to-one constraints, regularized by hybrid linear-nonlinear reconstruction decoder. The learned abundances are then used to construct an abundance affinity prior that guides a spectral transformer to emphasize class-discriminative interactions, and the resulting transformer features are fused with compact abundance descriptors for final prediction; in contrast to existing approaches that use abundance as auxiliary or concatenated features. Experiments on Indian Pines, Augsburg, and Berlin demonstrate the benefit of incorporating abundance- guided contextual modeling, particularly in heterogeneous urban scenes. The source code and trained models are available at: https://github.com/nnuvi/AGSA-Net

Continual learning (CL) enables models to acquire new knowledge from sequentially arriving tasks while retaining previously learned knowledge. However, in practical scenarios, task streams collected from untrusted sources may contain backdoor-poisoned samples, posing a critical challenge to the stability, plasticity, and security of continual learners. In this work, we investigate a challenging setting termed Continual Learning Under Backdoor Attack (CLUBA), where each incremental task may involve a small proportion of maliciously manipulated training samples. Unlike conventional continual learning or backdoor defense scenarios, CLUBA requires models to simultaneously mitigate catastrophic forgetting, preserve adaptation capability, and prevent the absorption of malicious supervision during sequential updates. To address this challenge, we propose a robust dynamic-expansion framework that integrates sample purification, selective recovery, and robust expert routing into a unified continual learning paradigm. Specifically, we introduce Bi-Prototype Purification (BPP) to identify suspicious samples by exploiting semantic discrepancies in feature space. Based on purified data, Gradient Discrepancy-based Robustness Optimization (GDBRO) selectively recovers informative poisoned samples through pseudo-label correction and gradient consistency evaluation, improving robustness while maintaining model plasticity. Furthermore, Robust Feature Consistency-based Expert Selection (RFCBES) constructs perturbation-aware class prototypes to enable reliable expert routing under corrupted or shifted inputs.

Linear algebra provides the framework of concepts (matrix rank, singular value decomposition (SVD), and eigendecomposition) that modern artificial intelligence employs to encode, compress, and propagate information through neural networks. This paper unifies fourteen separate peer-reviewed works analyzing the usage of these techniques in the context of transformer-based foundation model research, focusing on three areas of the topic: derivations and properties of self-attention matrices' output rank, compression methods that purposefully utilize this phenomenon, and the low-rank key-value (KV) cache projection and its semiseparable-matrix duality to linear attention and state-space structured models. We were motivated to conduct this work after observing an open problem in this literature: the interplay of the mentioned compression methods with natural rank collapse of the network. With this paper, we report an original finding that using SVD compression of attention projections actually has the opposite effect on the rank collapse of the network: while it strongly suppresses it at initialization, it accelerates on pretrained models (for GPT-2 124M, GPT-2 Medium 355M, and Pythia-160M) with minimal risk of object aliasing artifacts appearing (verified on all compression ratios) and is consistent across four rank estimation methods. A controlled causal decomposition of the effect in both settings showed that the reason for this behavior can be explained by the choice of the subspace SVD makes when compressing the matrix better than the reduction of the operator norm it achieves, explaining roughly 76% of the effect at initialization and 83% on the pretrained weights, providing a refinement to the calibration-aware compression viewpoint and an explanation of why it outperformed naive SVD truncation.

Accelerated MRI reconstruction recovers images from undersampled k-space. However, different acceleration factors produce distinct artifact patterns. Existing methods often train separate models for each factor, leading to poor cross-factor generalization and high training and storage costs. We propose Shared LoRA, a parameter-efficient framework that freezes the pretrained SHFormer backbone and trains a single shared set of LoRA adapters together with a lightweight gating network. During training, undersampled inputs are generated by randomly sampling acceleration factors and their corresponding sampling masks, enabling the shared adapters to learn reconstruction knowledge across factors. Given the acceleration factor, GateNet generates layer-wise coefficients to dynamically modulate the residual strength of each adapter. Experiments show that Shared LoRA achieves the best or competitive PSNR and SSIM across acceleration factors, while its trainable parameters account for only about 5.3% of the total model parameters. Its performance at lower acceleration factors remains largely unaffected as the jointly trained factor set expands, and it generalizes stably to unseen neighboring factors.

Aims. The objective of this work is to adapt a deep neural network model to perform galaxy morphological classification trained from crowd annotations, considering the training scheme, the agreement between the annotators, and the hierarchy. Methods. We use Galaxy Zoo 1 as our experimental testbed and trained a convolutional neural network (CNN) for the automatic classification of galaxies' morphologies. We analyze the impact of the following aspects on the classification accuracy and training efficiency: (i) Training only the last layer vs. training all the network; (ii) Classification with only the CNN vs. considering the hierarchy; (iii) Comparing the models trained with different amounts of data and levels of agreement between the annotators; (iv) Training by stages, transferring knowledge from one model to another; and (v) Combining several models as an ensemble. Results From the experiments, we derive the following results: (i) Training all the layers in the network significantly improves the accuracy (10% increase in exact match), compared to training only the last layer; (ii) There is a tradeoff between the amount of data and the level of agreement between the annotators used for training; (iii) Using the hierarchy can improve accuracy when the amount of training data is reduced; (iv) Training by stages through transfer learning (curriculum learning) produces higher accuracy for limited data; (v) Ensembles can improve accuracy; (vi) Models achieve a low accuracy for the most difficult cases, but, if we consider hierarchical measures, we can derive useful results for upper levels in the hierarchy. An accuracy above 99% is achieved when training all layers of the network and considering a high agreement between the annotators. Conclusions. Training deep learning models from crowd annotations involves additional challenges than learning from hard annotations.

The purpose of this research is to find data and methods using machine learning and deep learning to correctly predict the estimated travel time for transportation and logistics in a supply chain system. The supply chain ecosystem is very complex and heavily relies on the transportation and logistics of raw materials and finished goods. Accurate travel time estimation is critical because it helps supply chain members to improve logistics consistency and performance. This helps in planning, demand forecasting, lead time management and assembly planning. The logistics on the delivery side of the customer also plays a crucial role in customer satisfaction and voice of customer. With the collection of huge historical data and using novel techniques, the research builds an accurate model to predict travel time of inventory.

Search agents are usually trained under a single harness. But once an agent is deployed in a real application, its harness is frequently updated (e.g., a rewritten system prompt) to fit production needs. This exposes a fragility of post-trained agents: because a learned behavior is entangled with its training harness, even a harness update that leaves the task unchanged can fail to elicit the behavior. We train a search agent to perform parallel search, a popular strategy for improving both search efficiency and performance. We find that training under a fixed harness makes the behavior harness-local, overfit to that harness's surface form: when the harness changes, the model falls back to serial search. An intuitive fix is harness augmentation, but simply training on more harnesses does not resolve the problem. GRPO learns from the reward gap between parallel and serial rollouts of the same question: a small harness pool saturates that gap early, while a large pool dilutes the per-harness signal too thinly for any harness to consolidate. We therefore propose Curriculum HArness Rotation Training (CHART), a rotating curriculum that lets a search agent gradually consolidate parallel search across harnesses. At each periodic evaluation, CHART "graduates" the harnesses whose expected behavior is learned and replaces them with still-learnable ones, keeping the reward gap alive throughout training. Starting from the same harness pool, CHART makes the model learn parallel search on all harnesses, whereas static augmentation succeeds on at most half of them. The behavior also carries to held-out harnesses: CHART parallelizes on 89% of held-out turns, against at most 5% for the static pools. It further transfers to a new QA task and search environment, improving pass@1 by 5.6pp over the best static pool. Finally, CHART-trained agents benefit more from meta-harness search than baselines.