cs.LG · 2026-08-25 · No. 95

Machine Learning, 2026-08-25.

55 new papers in cs.LG. Titles, authors, abstracts. Links to arXiv. Want this in your inbox every morning? Subscribe →

01 — The papers

55 entries
  1. 01

    How to Train a Critic Stably and Efficiently

    Penghui Qi, Xiangxin Zhou, Wee Sun Lee

    cs.LG · cs.AI · cs.CL

    Group-based reinforcement learning methods such as GRPO for large language models avoid training a critic by sampling multiple responses for each prompt. A reliable critic could instead estimate token-level advantages from one response, but standard critic-based training recipes are often unstable. We study this instability and develop \textbf{Best-Practice Critic Optimization (BPCO)}, a recipe that combines DPPO, value predictions bounded to...

    arxiv.org/abs/2608.23566 · PDF

  2. 02

    Provably adaptive sampling with uniform and remasking discrete diffusion models

    Daniil Dmitriev, Zhihan Huang, Yuting Wei

    cs.LG · cs.IT · math.ST · stat.ML

    Discrete diffusion models offer a promising alternative to autoregressive generation by enabling parallel updates, but their sampling efficiency can depend strongly on the choice of the forward process and the sampler. For the uniform forward process, existing lower bounds for the standard $τ$-leaping sampler scale linearly with the ambient dimension $d$, raising the question of whether this dependence is intrinsic to the forward process. We...

    arxiv.org/abs/2608.23554 · PDF

  3. 03

    MetaCaster: Meta-Harness-Optimized Agent for End-to-End Few-Shot Learning of Lightweight Time Series Forecasters

    ChengAo Shen, Wenchao Yu, Fangyu Wu, Dongjin Song, Hanghang Tong, Dongsheng Luo, Wei Cheng, Haifeng Chen, Jingchao Ni

    cs.LG · cs.AI

    Time series forecasting (TSF) is evolving toward multimodal and agentic settings, yet using foundation models remains uneconomical in resource-constrained scenarios, where compact, specialized forecasters are more desirable. However, lightweight forecasters typically require substantial training data, limiting their use in domains with scarce, slowly accumulated, or privacy-sensitive time series. To address this dilemma, we investigate the...

    arxiv.org/abs/2608.23473 · PDF

  4. 04

    RAD: Rule-Augmented Relational Anomaly Detection

    Noah Dahle, Anne Tumlin, Ngoc Tran, Xenofon Koutsoukos, Tyler Derr

    cs.LG · cs.CR · cs.DB

    Anomaly detection is often applied to data stored in relational databases, yet most existing methods require flattening multiple tables into a single feature matrix. This flattening can obscure entity identity, schema structure, and multi-hop dependencies, limiting the detection of anomalies that depend on relational context rather than isolated feature values. Beyond preserving relational structure, relational anomaly detection raises an...

    arxiv.org/abs/2608.23468 · PDF

  5. 05

    ProxyFormer: A Dual-Stream Proxy Architecture for Ultra-Long Context and High-Resolution Generation

    Zhongpan Tang

    cs.LG

    The quadratic growth of attention computation and key-value (KV) cache with respect to sequence length is a central bottleneck for ultra-long-context language models and high-resolution generative models. We propose ProxyFormer, a general dual-stream architecture built upon proxy tokens. In each layer, fine-grained local features are compressed bottom-up into a small set of proxy states; expensive global interactions are performed only in the...

    arxiv.org/abs/2608.23463 · PDF

  6. 06

    Diversity-Based Active Learning: An Evaluation of Metric Spaces for Active Learning Selection

    Siddharth Chilamkur, Dorit S. Hochbaum

    cs.LG

    With rapid advancement over the last few years, many different methods are now widely used for classification. However, training these models requires substantial labeled data. Active Learning is a potential solution to this problem. Pool-based active learning minimizes costs by querying only the most informative samples from an unlabeled dataset. Diversity-based approaches, on the other hand, attempt to select a representative subset of the...

    arxiv.org/abs/2608.23461 · PDF

  7. 07

    Traceable Spectral Inference via Influence Functions: Efficient Data Attribution and Error Proxies for the Ariel Mission

    Nikki Grens, Luís F. Simões, Kai Hou Yip, Theresa Lueftinger

    cs.LG · astro-ph.IM · stat.ML

    Interpretability is critical for machine learning models deployed in scientific space missions such as ESA's Ariel, where ground truth is unavailable during operations and physical plausibility must be assessed. While most explainable AI methods focus on feature attribution, this work investigates training data attribution through influence functions and introduces three key contributions for operational spectroscopy pipelines. First,...

    arxiv.org/abs/2608.23458 · PDF

  8. 08

    ChebBooster: A Training-Free Approach for Efficient Diffusion Transformer Inference via Chebyshev-Inspired Extrapolation

    Chengjie Lu, Tianchi Deng, Zhengqi He, Chengwen Luo, Xueliang Li

    cs.LG · cs.AI

    Diffusion Transformers (DiTs) have shown strong performance in high-fidelity image generation, but their sampling process remains computationally intensive due to full model execution at every timestep. While cache-based acceleration has been explored to mitigate inference cost, naive reuse schemes suffer from low accuracy over long intervals, and Taylor-series-based extrapolation methods often face instability caused by Runge oscillations....

    arxiv.org/abs/2608.23429 · PDF

  9. 09

    The Axiomatic Trader: Latent Regularity, Information Budgets, and the Canonical Form of a Quantitative Investment System

    Jiayu Li

    cs.LG · q-fin.PM

    Systematic trading rests on one article of faith: that regularities found in the past persist. We state it as a time-invariant mechanism driven by an unobserved latent state, and show that it leaves a researcher five constants to declare --- the recurrence bound $Lambda$ at a block length $b$, the invariance defect $epsilon_0$ of the representation it is declared of, the coherence times $ell_i$ of the state's coordinates, the signal ceiling...

    arxiv.org/abs/2608.23416 · PDF

  10. 10

    Spectrum-Aware Bounds on Invertibility for Privacy-Enhancing Instance Encoding

    Seokjin Hwang, Yuting, Li, Kiwan Maeng

    cs.LG · cs.CR

    Instance encoding is a popular empirical technique for privacy enhancement when sharing data to an untrusted server. It transforms sensitive data through an encoding process before sharing, with the hope that the encoding process retains utility but makes it hard to reconstruct the original data. However, most work offers no theoretical guarantee that the encoding process is actually irreversible. A recent work derived a mean-squared error...

    arxiv.org/abs/2608.23382 · PDF

  11. 11

    Test-Time Adaptation for ECG Classification via SQI-Gated Self-Training and Beat-Rhythm Consistency

    Wenhan Jiang, Zhipeng Deng, Jiale Zhou, Haolin Wang, Yafei Ou, Yefeng Zheng

    cs.LG

    Deep learning models for electrocardiogram (ECG) classification often suffer from significant performance degradation when deployed in unseen domains due to shifts in acquisition devices and patient populations. Test-time adaptation (TTA) offers a practical solution by adapting models using only unlabeled data at inference time. However, existing TTA methods often underperform on ECG tasks, since naive online updates ignore the hierarchical...

    arxiv.org/abs/2608.23347 · PDF

  12. 12

    Towards Actionable Surgical Team Dynamics: from Teamwork to Counterfactual Annotations

    Vincenzo Marco De Luca, Antonio Longa, Andrea Passerini

    cs.LG

    Modeling team interactions in high-stakes environments such as operating rooms is critical for understanding how coordination, communication, and individual behaviors shape team performance and safety outcomes. Existing datasets in this domain are often fragmented across modalities, annotation schemes, and formats, limiting their ability to support integrated analyses of real-world collaborative processes. We address this limitation by...

    arxiv.org/abs/2608.23344 · PDF

  13. 13

    Beyond Point Predictions: Uncertainty-Aware Satellite Poverty Mapping for Public Policy

    Markus B. Pettersson, James Bailie, Mohammad Kakooei, Eagon Meng, Adel Daoud

    cs.LG

    Despite their critical importance for policy and research, high-resolution poverty data remain limited across much of Africa. Machine learning (ML) with earth observation (EO) imagery has recently emerged as a way to supplement these data by predicting (i.e., estimating) poverty where it has not been directly measured. Yet to be used reliably, decision-makers and analysts need assurances that they will not be misled by the errors in these...

    arxiv.org/abs/2608.23322 · PDF

  14. 14

    Sigmoid Attention as a Better Substrate for Learned KV Cache Eviction

    Isaac, Li

    cs.LG · cs.AI

    Learned KV-cache eviction often faces a soft-to-hard mismatch: during training, differentiable gates typically attenuate token contributions, whereas inference saves memory only when KV entries are physically removed. We ask whether the attention substrate affects this soft-to-hard transition. Using GPT-2-scale Transformers trained on OpenWebText, we run a controlled $2\times2\times2$ comparison over attention type, learned gating, and...

    arxiv.org/abs/2608.23296 · PDF

  15. 15

    Poisson Subspace Clustering: Focusing on the Essentials in Count Data

    Collin Leiber, Kai Puolamäki, Heikki Mannila

    cs.LG

    Count data represented as a matrix of non-negative integer values, such as contingency tables, are prevalent across diverse domains. When clustering such data sets, specific methods are required, as generic algorithms often fail to consider their unique distributional properties, leading to unreliable outputs. An effective strategy is to use well-established statistical models such as the Poisson and negative binomial distributions. We...

    arxiv.org/abs/2608.23287 · PDF

  16. 16

    How Much Regularization Survives Averaging? Update Masking in Federated Learning

    Wenhao Yan, Fu Kuroda, Yucheng Jin, Zhenke Chen

    cs.LG · cs.AI

    Federated learning on non-IID data seeks flat minima to generalize across clients, and existing methods borrow sharpness-aware minimization from centralized training. There is a second way to reach flat minima, in which the regularization comes for free from noise added to the parameter updates, and it has never been carried over to the federated setting. We show the reason. Masking charges the optimizer for moving in sharp directions. We...

    arxiv.org/abs/2608.23286 · PDF

  17. 17

    A Multidimensional Data-Driven Hybrid Transformer Framework for Non-invasive Continuous Blood Pressure Prediction

    Yuexin Ma, Jingqi Hou, Yuxuan Kang, Zhaoying Liu

    cs.LG

    Objective. To develop and evaluate a cuffless continuous blood pressure (BP) estimator using temporal physiological and demographic features. We propose a hybrid Transformer framework to estimate diastolic and systolic BP from ECG/PPG-derived feature sequences. Approach. Rather than raw waveforms, the framework models 10-step sequences of six physiological descriptors and two demographic covariates. A Multi-Source Temporal Encoder Module...

    arxiv.org/abs/2608.23276 · PDF

  18. 18

    From Multimodal Observation to Interpretable Suggestions: Counterfactual Time-Expanded Relational Modeling of Surgical Teams

    Vincenzo Marco De Luca, Antonio Longa, Giovanna Varni, Andrea Passerini

    cs.LG

    In surgery, patient safety is threatened not only by technical issues but also by poor teamwork. However, existing surgical AI-based solutions focus mainly on visual workflow and technical execution, neglecting the modeling of team interactions and missing opportunities to actively support clinicians in improving their teamwork skills. To address this gap, we propose a tempo-relational framework for modeling surgical team dynamics from...

    arxiv.org/abs/2608.23254 · PDF

  19. 19

    The Laws of Context Allocation: Causal Measurement and Closed-Loop Orchestration in Generative Search

    Peiyang Liu, Xi Wang, Di Liang, Wei Ye

    cs.LG · cs.CL · cs.IR

    As Retrieval-Augmented Generation (RAG) shifts toward diverse portfolio generation, it is stymied by two critical bottlenecks: flawed measurement of evidence utilization, and suboptimal context budget allocation. We resolve both sequentially. To resolve measurement, we expose a pervasive ``diagnostic illusion'': standard relevance proxies fail catastrophically on hard negatives. We replace them with an efficient causal leave-one-out probe...

    arxiv.org/abs/2608.23252 · PDF

  20. 20

    A Comparative Study of Label-free Representation Quality Metrics in Deep Learning

    Daniel Richards Arputharaj, Daniel Jönsson, Gabriel Eilertsen

    cs.LG · cs.CV

    We present a comparative study of label-free metrics for assessing the quality of representations in deep neural networks to understand their reliability under a wide variety of configurations. We group existing label-free metrics into three families based on their construction and analytically establish connections between metrics within the same family. We then characterise the sensitivity of spectral metrics through controlled synthetic...

    arxiv.org/abs/2608.23182 · PDF

  21. 21

    Counterfactual Transition Graphs: Evaluating Cross-Class Transition Quality

    Syed Muhammad Hamza Zaidi, Szymon Bobek, Grzegorz J. Nalepa, Myra Spiliopoulou

    cs.LG · cs.AI

    Counterfactual (CF) explanations for time-series classifiers are usually evaluated one example at a time: what minimal edit flips this single window's prediction? We argue that the more informative question for diagnostic interpretability is structural: how does the classifier connect its own classes to each other? We propose a counterfactual transition graph (CGT) in which each node is a class and each edge weight is the CF reliability of...

    arxiv.org/abs/2608.23164 · PDF

  22. 22

    When More Modalities Hurt: Modality Dropout for Heavy-Duty Vehicle Engine Diagnostics

    Adeel Zafar, Slawomir Nowaczyk, Hamid Sarmadi, Saeed Gholami Shahbandi

    cs.LG

    Heavy-duty vehicle diagnostics generate three disconnected data modalities: unstructured multi- lingual service complaints, high-dimensional sensor telemetry with over 80% missing values, and Diagnostic Trouble Codes (DTCs). We investigate whether fusing these modalities improves engine component classification on a proprietary dataset from a major truck manufacturer. Through 5-fold cross-validation across multiple model configurations...

    arxiv.org/abs/2608.23161 · PDF

  23. 23

    Conformal Risk Minimization for Semi-Supervised Domain Adaptation via Optimal Transport

    Manos Giannopoulos, Yi Shen, Michael M. Zavlanos

    cs.LG

    In high-stakes healthcare applications, machine learning models are frequently trained on data from one patient population and deployed on another, creating a distribution shift that degrades both accuracy and reliability. Semi-Supervised Domain Adaptation (SSDA) addresses this by leveraging labeled data from some source domain to improve model performance on a target domain where labels are scarce. However, existing SSDA methods optimize...

    arxiv.org/abs/2608.23153 · PDF

  24. 24

    Activation-Weighted Seeded Residual Coding for Low-Bit LLM Weight Repair

    Zehao Liu, Chuangchuang Fang, Yang Ren

    cs.LG · cs.CL

    Low-bit weight quantization saves storage but leaves errors that degrade language-model quality. We introduce Activation-Weighted Seeded Residual Coding (AWSRC), a compact repair codec for an existing quantization backbone. Given a reconstructed weight $W_0$, AWSRC encodes the residual $W-W_0$ using deterministic seed-generated bases. The sidecar stores seed selectors, low-bit coefficients, and scales rather than an explicit codebook....

    arxiv.org/abs/2608.23144 · PDF

  25. 25

    DeMixPert: Decomposed Response Modeling with Gaussian Mixtures for OOD Single-Cell Perturbation Prediction

    Jiawen Liu, Xuechenxiao Cao, Yutong Li, Bing Liu, Jiaming Liang, Tinghe Zhang, Xiaoqi Sheng, Hongmin Cai

    cs.LG · cs.AI

    Predicting transcriptome-wide responses to unseen genetic perturbations remains a major computational challenge because accurate prediction requires recovering both perturbation-specific transcriptional shifts and heterogeneous cellular responses. Existing methods often entangle deterministic response structure with stochastic population-level variation, causing dominant shared patterns to mask weaker perturbation-specific signals and impair...

    arxiv.org/abs/2608.23114 · PDF

  26. 26

    PolyChirp: Multi-Species Birdsong Classification Using TinyML on Low-Power Acoustic Sensors

    Nathan Duboisset, Zhaolan Huang, Felix Bießmann, Roudy Dagher, Antoine Lavandier, Emmanuel Baccelli

    cs.LG · cs.AI

    Recent progress in the field of TinyML has demonstrated that low-power hardware based on microcontrollers can achieve bird species monitoring in real time based on acoustic sensor data for an entire breeding period on a single battery charge. However, the state of the art on low-power microcontrollers was so far limited to binary classification of a single species. In contrast, real fauna monitoring deployments often target multiple species...

    arxiv.org/abs/2608.23101 · PDF

  27. 27

    Macro-Action Topological Navigation under Noisy Localization using Reinforcement Learning

    Simon Hakenes, Tobias Glasmachers

    cs.LG · cs.RO

    Navigating large, photorealistic 3D apartments from raw pixels is widely considered infeasible for plain reinforcement learning. We build an agent that does it anyway, estimating its own pose from the camera alone. The agent has to reach several target objects in sequence, and their positions change between episodes, so it must explore to find them. It builds on our earlier object-centric topological controller, which still read the agent's...

    arxiv.org/abs/2608.23055 · PDF

  28. 28

    Graph Representation Learning of Lightweight IoT Ciphers

    Jonathan Cook, Sabih ur Rehman, M. Arif Khan

    cs.LG · cs.CR

    SIMON and SIMECK belong to a family of Lightweight Cryptographic Algorithms (LCAs) based on the Feistel block cipher, designed for Internet of Things (IoT) devices. As with all Feistel ciphers, they are susceptible to differential cryptanalysis, necessitating rigorous resilience evaluations. While state-of-the-art techniques leverage heuristics and sampling to improve efficiency, little work has applied Machine Learning (ML) guided Graph...

    arxiv.org/abs/2608.23054 · PDF

  29. 29

    Reservoir of Importance: Learning Semi-Structured Sparsity with Differentiable Subset Sampling

    Ha Dinh, Xuan Duy Ta, Khoat Than, Khac-Hoai Nam Bui

    cs.LG

    Semi-structured $N$:$M$ sparsity has emerged as a practical direction for accelerating large language models (LLMs). However, existing learnable-mask approaches incur substantial parameter and memory overhead, limiting their scalability to large models and aggressive sparsity regimes. In this work, we revisit semi-structured pruning from a perspective that reconciles efficiency with scalability. We propose Reservoir of Importance (RoI), a...

    arxiv.org/abs/2608.23048 · PDF

  30. 30

    ST$^2$U: Stateful Test-Time Unlearning via Restricted Knowledge Boundary Control

    Xunlei Chen, Qinghui Gong, Ruini Xue, Yaodong Hu, Tian Lan, Wenhong Tian

    cs.LG · cs.CL

    Controlling restricted knowledge in large language models is essential for model alignment and safe deployment. Test-time unlearning avoids costly retraining and parameter updates by intervening only during inference. However, existing activation-editing methods apply isolated pointwise corrections, overlooking how autoregressive generation continually reconstructs hidden states from the prompt, cache, and generated prefix. Consequently,...

    arxiv.org/abs/2608.23034 · PDF

  31. 31

    FedCC: Towards Addressing Label Distribution Skews in Distillation-Based Federated Learning

    Wenxuan Ye, Onur Ayan, Xueli An, Georg Carle

    cs.LG · cs.AI

    Federated Learning (FL) enables distributed clients to collaboratively train models without sharing raw data, making it promising for leveraging massive devices in communication networks. In distillation-based FL, each client applies its local model on an unlabeled public dataset, and shares only prediction results with the server. While heterogeneous local data introduces label distribution skew, thus biasing client models toward majority...

    arxiv.org/abs/2608.23031 · PDF

  32. 32

    SplitLite: Low-Rank Residual Compression for Split Learning

    Tao Li, Yulin Tang, Qi Guo, Xianhao Chen

    cs.LG · cs.AI

    Federated fine-tuning of on-device large language models (LLMs) faces a significant computing burden. To overcome this limitation, split learning (SL) has emerged as a promising solution, which offloads the primary training workload to a powerful server. However, SL requires exchanging high-dimensional activations and gradients between clients and the server, resulting in prohibitive communication costs. To overcome this challenge, we propose...

    arxiv.org/abs/2608.23018 · PDF

  33. 33

    A Physical Response-and-Memory Model for Muon Optimization

    Yinze Hu, Hongjun Xiang, Xingao Gong, Hongyu Yu

    cs.LG · cond-mat.dis-nn · cond-mat.stat-mech · cs.AI · physics.comp-ph

    Training large language models is costly. How low a loss the same compute can ultimately reach depends on how each step's gradient is converted into a weight update; the rule that performs this conversion is the optimizer. From SGD and AdamW to the recent Muon, effective update rules have mostly been shaped by engineering intuition and then selected on benchmarks. Muon semi-orthogonalizes the momentum matrix before applying the update and has...

    arxiv.org/abs/2608.22994 · PDF

  34. 34

    Hierarchy-Aware Semantic Losses for Knowledge Graph Link Prediction

    Filip Kronström, Ross D. King

    cs.LG

    Knowledge graphs are often accompanied by ontological class hierarchies that encode valuable semantic information, yet many link prediction methods either ignore such hierarchies or incorporate them indirectly through additional graph edges. Recent work introduced hierarchy-aware graph neural networks (GNNs), which use semantic losses derived from box embeddings to encourage satisfaction of subclass relationships during GNN-based...

    arxiv.org/abs/2608.22981 · PDF

  35. 35

    Do Time-Series Foundation Models Pay Off for Industrial Monitoring? A Cost-Aware Empirical Study

    Guan-Hua Wen, Kuan-Yu Chen

    cs.LG

    Industrial monitoring models must detect operationally relevant deviations while satisfying target-specific data, calibration, and resource constraints. Time-series foundation models (TSFMs) promise reusable representations and zero-shot forecasts, yet evidence for their deployment value remains mixed when task definitions are heterogeneous and lightweight baselines are competitive. This work presents a protocol-aware empirical assessment...

    arxiv.org/abs/2608.22968 · PDF

  36. 36

    Stochastic gradient descent with initial regularization

    Nabil Kahalé

    cs.LG · math.OC · stat.ML

    We analyze a variant of stochastic gradient descent with initial regularization (SGDIR) and derive dimension-free upper bounds on its expected excess risk for the squared loss. In the noiseless case, we obtain new bounds for both averaged and non-averaged SGDIR under moment, source, and capacity assumptions. For a particular value of the source parameter, these bounds are of order $m^{-2}\log^{2}m$, where the number of training samples is of...

    arxiv.org/abs/2608.22953 · PDF

  37. 37

    A Momentum-Based Variance-Reduced Algorithm for Federated Multiobjective Optimization

    Yong Zhao, Chunlin You, Minh N. Dao, Zai-Yun Peng

    cs.LG · math.OC

    Federated learning has traditionally been formulated as a single-objective optimization problem, primarily focused on maximizing model utility. In real-world applications, however, machine learning models often need to optimize multiple and potentially conflicting objectives simultaneously. This motivates federated multiobjective optimization (FMOO), which provides a natural framework for jointly handling multiple task-specific objectives in...

    arxiv.org/abs/2608.22945 · PDF

  38. 38

    Safety Hacking in Constrained Best-of-$N$ Inference-time Scaling

    Akifumi Wachi, Takumi Tanabe, Youhei Akimoto

    cs.LG · cs.AI · cs.CL · cs.CR

    Inference-time pipelines often sample multiple outputs, filter them with a learned safety model, and return the proxy-feasible output with the highest learned reward. We show that this composition creates a two-stage failure: an imperfect safety proxy first contaminates the feasible set with unsafe outputs, and reward maximization can then amplify this residual contamination. We define \emph{safety hacking} as selecting an output that passes...

    arxiv.org/abs/2608.22915 · PDF

  39. 39

    The Mask Is Not the Model: Auditing Prefix Invariance in Attention, State-Space, and Hybrid Sequence Models

    Taebong Kim, Youngsik Hong, Minsik Kim, Sunyoung Choi, Jaewon Jang, Minseo Kim

    cs.LG · cs.AI

    We formalize prefix invariance: representations at position t must not depend on future inputs. We give a lightweight audit, two forward passes, no training or gradients, that localizes exactly where causality breaks. Attention-mask inspection is incomplete: leaks can occur via scans or normalization despite correct masks. Across 192 injected-fault trials on eight checkpoints, mask inspection found none, while our audit localized all 192/192,...

    arxiv.org/abs/2608.22876 · PDF

  40. 40

    Stochastic Separability of Embedding Manifolds

    Liqing Zhang

    cs.LG · cs.CV

    Neurobiological studies and representation learning have observed that representations of objects belonging to the same category in high-dimensional neural spaces exhibit low-dimensional object manifold characteristics, and different object manifolds are linearly separable in these neural spaces. However, these experimentally observed phenomena lack rigorous theoretical validation to date. This paper proposes a new stochastic separability...

    arxiv.org/abs/2608.22874 · PDF

  41. 41

    Mapping the Concept Landscape: Structural Perception of Global Distributions for Transparent Data Pruning

    Dongyue Wu, Tao Ma

    cs.LG · cs.CV

    Existing data pruning methods predominantly rely on high-dimensional feature embeddings to measure sample importance. However, these compressed vectors often obscure fine-grained semantic interactions, leading to suboptimal coverage of rare semantic concepts in the pruned subsets. In this paper, we propose Mapping the Concept Landscape (MCL), a novel structural perception framework for transparent data pruning. Instead of abstract embeddings,...

    arxiv.org/abs/2608.22858 · PDF

  42. 42

    Thinking at the Right Size: Amortized Distillation Across Post-Trained LLMs

    Yan Zhou, Sara Kangaslahti, Jonathan Geuter, Nihal V. Nayak, Marco Fumero, Francesco Locatello, David Alvarez-Melis

    cs.LG

    Practical deployment of large language models (LLMs) requires families of post-trained variants---instruction-tuned, reasoning-tuned, and chat-style models---each at multiple sizes to meet diverse latency and memory budgets. Producing each (variant, size) pair independently is prohibitive, so model families typically span only a handful of coarse-grained sizes per post-trained variant. Boomerang distillation (Kangaslahti et al., 2026) reduces...

    arxiv.org/abs/2608.22854 · PDF

  43. 43

    RIBOSPAN: A Long-Context RNA Foundation Model for Versatile RNA Modeling

    Ziyuan Wang, Bohao Tang, Fei Zhang, Shuo Han, Pengfei Liu

    cs.LG · q-bio.GN

    Full-length RNAs, particularly messenger RNAs, often exceed the context lengths used to pretrain existing RNA foundation models, limiting complete-transcript modeling at single-nucleotide resolution. We present RIBOSPAN, a 1.61-billion-parameter bidirectional RNA foundation model natively pretrained with context lengths up to 10,240 nt. RIBOSPAN combines dense bidirectional self-attention, single-nucleotide tokenization, and...

    arxiv.org/abs/2608.22849 · PDF

  44. 44

    Hierarchy-Aware Supervised Uncertainty Estimation for Black-box LLM Taxonomic Reasoning

    Shuting Xie, Nathaniel Lesperance, Graham W. Taylor

    cs.LG · cs.AI

    Large language models (LLMs) are increasingly used for scientific decision support, yet reliable confidence estimation remains difficult in black-box settings. We study uncertainty estimation for hierarchical taxonomic reasoning generated by a black-box LLM in a long-tailed biodiversity monitoring pipeline. Using proxy features extracted by an open-source tool LLM, we train lightweight supervised estimators with hierarchy-aware supervision to...

    arxiv.org/abs/2608.22839 · PDF

  45. 45

    DIME: Query-Efficient Framework for Membership Inference on Diffusion Models

    Tue Do, Daniel Alabi

    cs.LG · cs.CR

    Membership inference attacks expose whether individual records were used to train a model, yet existing attacks on diffusion models are largely heuristic and can require substantial query budgets. We introduce DIME (Denoiser Ideal Membership Error), a theoretically grounded and query-efficient framework for membership inference on diffusion models. Our starting point is an exact characterization of the optimal diffusion denoiser for a finite...

    arxiv.org/abs/2608.22824 · PDF

  46. 46

    Fairness-Aware Mixture-of-Experts via Subgroup Reweighting and Gate Regularization

    Sunhee Hwang

    cs.LG · cs.AI

    Deep learning models often produce performance disparities across demographic groups, due to the training data imbalance with respect to sensitive attributes such as gender or age. To address this problem, existing work has explored fair representation learning, data re-sampling, and adversarial training, which can be broadly categorized into two main approaches. Single-stage methods typically learn a shared representation for fairness, but...

    arxiv.org/abs/2608.22820 · PDF

  47. 47

    SAGE: Stability-Aware Graph-Based Ensemble Feature Selection for Explainable Postpartum Depression Risk Prediction

    Md. Rokon Islam Emon, Syed Shariar Alam Shuvo, Shahriar Siddique Ayon, Abdullah Al Mamun, Ahnaf Atef Choudhury

    cs.LG

    Postpartum depression (PPD) poses a major burden on maternal and child health, especially in low- and middle-income countries where prevalence exceeds 19%. Despite advancements in machine learning for PPD prediction, current approaches are limited by opaque global explanations that lack clinical usefulness at the patient level, unstable feature selection, and poor generalization under class imbalance. We propose SAGE, a Stability-Aware...

    arxiv.org/abs/2608.22809 · PDF

  48. 48

    CatchBench: When Can an Agent Failure Be Caught?

    Yue Zhao

    cs.LG · cs.MA · cs.PF

    When can an agent failure be caught? An audit is usually limited by the record rather than by the method. CatchBench therefore puts one auditor's question to three information states: the declared configuration before a run (PRE), a growing prefix of its trace (LIVE), and the finished trace (POST). Prior benchmarks fix one of these states or vary the telemetry; to our knowledge none scores all three under one task-method interface. Each state...

    arxiv.org/abs/2608.22808 · PDF

  49. 49

    Change Detection in Probability Flow ODE: Online Testing in Diffusion Latent Spaces

    Artem Kraevskiy, Artem Prokhorov

    cs.LG · stat.ML

    A rapidly growing range of sequential data tasks, such as identifying trend reversals in financial markets, auto-segmenting video and audio recordings, detecting changes in movement direction from motion sensors cannot be fully addressed without detection of distributional shifts in time-ordered data. We consider a sequential change-point detection problem where the conditional density switches at an unknown time, yet neither the pre- nor...

    arxiv.org/abs/2608.22807 · PDF

  50. 50

    Contrastive Representation-Guided Genetic Minority Oversampling for Imbalanced Time-Series Classification

    Wenbin Pei, Yunrong Hao, Zhen Liu, Guan Wang, Bing Xue, Yiu-Ming Cheung, Qiang Zhang

    cs.LG

    Real-world time-series classification tasks often exhibit class imbalance, which can be extremely severe in some applications. To avoid training biased classifiers on imbalanced data, sampling is one of the most popular data pre-processing techniques because of its classifier-agnostic nature. However, due to the complex temporal dependencies in original time-series data and the scarcity of minority-class samples, existing sampling methods,...

    arxiv.org/abs/2608.22804 · PDF

  51. 51

    ReCoG: Reciprocal Co-Evolution for Multimodal Graph Learning

    Rui Xue, Tianfu Wu

    cs.LG

    Multimodal graph learning requires jointly training over graph structure and heterogeneous node attributes, yet existing methods largely decouple these processes: prior multimodal graph neural networks (GNNs) focus on aligning modalities in a shared embedding space while operating on fixed or weakly adapted graph structures, and graph structure learning approaches infer topology from unimodal node representations without accounting for...

    arxiv.org/abs/2608.22786 · PDF

  52. 52

    Neural Operator based Multi-Field Reconstruction of Inner Solar Boundary State

    Vignesh Kumar Pandian Sathia, Reza Mansouri, Dustin J. Kempton, Pete Riley, Rafal A. Angryk

    cs.LG · astro-ph.IM · astro-ph.SR · cs.CV

    The Solar wind is a continuous flow of charged particles emanating from the solar surface and governed by complex, interacting magnetohydrodynamic processes. Accurate specification of inner-boundary conditions is essential for heliospheric modeling and solar-wind prediction. In many practical applications, only a subset of interacting multi-field variables is directly available, but for a comprehensive view of solar wind prediction and...

    arxiv.org/abs/2608.22782 · PDF

  53. 53

    Learning to Control Coupled-Dynamics Environments with Joint Markov Decision Processes

    Ege C. Kaya, Aliasghar Pourghani, Mahsa Ghasemi, Vijay Gupta, Abolfazl Hashemi

    cs.LG · math.OC

    Coupled-dynamics environments expose the one-step outcomes that would follow from several possible counterfactual actions under a common realization of exogenous randomness. The ordinary Markov decision process formalism allows one to reason about the marginal law of each action but discards dependence across these counterfactual outcomes. The Joint Markov decision process (JMDP) formalism preserves that dependence. Prior work established the...

    arxiv.org/abs/2608.22765 · PDF

  54. 54

    LpWM: A Case for Sparse Representations in World Models

    Yilun Kuang, Yash Dagade, Quentin Le Lidec, Lucas Maes, Randall Balestriero, Yann LeCun

    cs.LG

    Joint-embedding predictive architectures (JEPAs) learn latent dynamics for planning and avoid representation collapse by matching features to maximum-entropy distributions such as isotropic Gaussians, yielding dense representations. However, it is unclear whether dense representations are the most favorable geometry for modeling dynamics. In this work, we ask whether a different geometry, sparse representations, can make action-conditioned...

    arxiv.org/abs/2608.22764 · PDF

  55. 55

    MOSH-WM: Mask-Grounded Soft-Hamiltonian Dynamics for Object-Centric World Models

    Zhekai Wang, Haoxiang Huang, Xiang Liu, Zhikang Chen, Yueqing Sun, Qi Gu, Shiji Zhou, Miao Liu, Sen Cui

    cs.LG

    Object-centric world models forecast future videos by evolving a set of entity slots, but the variables receiving dynamics supervision are often unconstrained visual features. We introduce \method{}, a mask-grounded soft-Hamiltonian world model that makes its position-like state explicitly depend on slot-owned image support. A frozen video-slot encoder produces slots and masks; spatial moments of mask-owned support form a canonical state $Q$,...

    arxiv.org/abs/2608.22750 · PDF

This edition is part of The Daily Abstract — cs.LG archive. Subscribe to receive these in your inbox each morning, automatically translated to Spanish, with reply-to-PDF: arxivdaily.ignorelist.com.

Colophon Set in Georgia, with system sans for interface chrome and a monospaced stack for code and paper identifiers. Sole accent: amber #D99C5E. Built and served on an always-free VM. The masthead is set 14% letterspaced because newspapers do that and it works.