cs.LG · 2026-10-07 · No. 136

Machine Learning, 2026-10-07.

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

01 — The papers

60 entries
  1. 01

    Conformal Prediction Sets Quantify Information Gain: A Theoretical Perspective

    Kevin Zhang, Stephen Bates

    cs.LG

    Conformal prediction is a popular tool for uncertainty quantification that outputs prediction sets with finite-sample coverage guarantees. While prediction set size is commonly used as a heuristic measure of uncertainty, the information-theoretic basis for this interpretation remains poorly understood. In this work, we provide such a foundation using a decision-theoretic generalization of entropy tailored to set-valued prediction. In...

    arxiv.org/abs/2610.08785 · PDF

  2. 02

    Neural Petri flows for chemical reactions

    Jose Eduardo Escrig Molina, Daniel Probst

    cs.LG · physics.chem-ph · q-bio.QM

    Petri nets have been used to describe chemical processes such as reactions.They map well to chemistry: Places are the bonds between atoms and the free valence of each atom, a token is a unit of bond order, a transition forms or breaks a bond, the conserved quantities are the valence budgets of the atoms, and the enabling rule is the valence rule. These semantics are not guaranteed by learned models of reactions or neural networks that are...

    arxiv.org/abs/2610.08750 · PDF

  3. 03

    Linear Bandits under Exact Sliding-Window Constraints

    Seyed Mohammad Hadi Hosseini, Yasin Abbasi-Yadkori, Sattar Vakili

    cs.LG

    We study linear bandits under exact sliding-window constraints, where every consecutive block of actions must belong to a prescribed feasible set. In the offline setting, where the reward function is known, we show that convexity and cyclic-shift invariance make a stationary solution optimal when $w\mid T$ and within an additive $O(w)$ gap otherwise. In the online setting, we show that geometric structure alone is insufficient for learning,...

    arxiv.org/abs/2610.08745 · PDF

  4. 04

    Reinforcement Learning with Conformal Action Sets: An Application to Sequential Recommendation

    Wenwen Si, Honghao Wei

    cs.LG · cs.AI

    Sequential recommenders typically use a fixed slate size even though the number of useful alternatives changes within a session. We propose Reinforcement Learning with Calibrated Pruning (RLCP), which adapts the retained action set using critic scores and an online threshold. The threshold is updated from binary feedback indicating whether the set contains an action in a proxy target. We prove a deterministic bound on the observed proxy miss...

    arxiv.org/abs/2610.08743 · PDF

  5. 05

    On the Computational Tractability of Robust Bandits

    Vanessa Kosoy, Vinayak Pathak

    cs.LG

    Learning when the environment does not belong to the learner's hypothesis class is typically handled using agnostic learning guarantees. However, for anything beyond supervised learning, agnostic guarantees are difficult to come by. Recently, imprecise bandits (Kosoy, 2025) (later renamed to robust bandits in Appel and Kosoy, 2025) were introduced as another approach to unrealizable learning in the bandits setting and a $Θ(\sqrt{T})$ regret...

    arxiv.org/abs/2610.08740 · PDF

  6. 06

    Optimal and Efficient Online Inverse Optimization

    Anupam Gupta, Guru Guruganesh, Honghao Lin, Vahab Mirrokni, Renato Paes Leme, David P. Woodruff

    cs.LG · cs.DS

    In online inverse linear optimization, a learner recommends an action and then observes the choice of an expert who maximizes a fixed, unknown linear objective on $\mathbb{R}^{d}$; the goal is to learn to optimize this objective without observing it. Sakaue recently obtained the optimal regret $O(\sqrt d)$ with a randomized algorithm making $(dT)^{O(d)}$ linear optimizations per round, and asked whether it can be attained in polynomial time....

    arxiv.org/abs/2610.08735 · PDF

  7. 07

    GeneICL: A Tabular Foundation Model for Bulk Transcriptomics

    Michael Bohl, Alexander Theus, David Wissel, Valentina Boeva

    cs.LG

    Gene expression is widely measured in biomedicine, yet clinical outcome prediction remains challenging due to high dimensionality, strong feature correlations, and limited labeled data. Large self-supervised transcriptomic foundation models often fail to outperform simple supervised baselines. Tabular foundation models offer an alternative through in-context learning, but are typically pretrained on generic synthetic data rather than...

    arxiv.org/abs/2610.08694 · PDF

  8. 08

    Probabilistic Counterfactual Inference for Discrete Outcomes in Gaussian-Process Causal Models

    Juliette Sinnott, Amir-Hossein Karimi, Mohammad Kohandel

    cs.LG

    Counterfactual inference in Gaussian-process structural causal models (GP-SCMs) has been developed primarily for continuous endogenous variables, limiting applicability to causal graphs that contain discrete child nodes with continuous parents. We introduce a unified probabilistic framework for counterfactual inference with heterogeneous variable types by pairing GP predictors with explicit exogenous noise mechanisms. For discrete outcomes,...

    arxiv.org/abs/2610.08689 · PDF

  9. 09

    A Systematic Study of Small Language Models on Abstract Reasoning Tasks

    Nur A Zarin Nishat, Jens Lehmann, Andrei Aioanei, Sahar Vahdati

    cs.LG · cs.CL

    Endpoint accuracy on abstract-reasoning benchmarks does not reveal whether a language model has acquired a transferable rule or fit distribution-specific regularities. We study this distinction in small language models on the ARC-TGI benchmark, which organizes abstract grid transformations into controllable task families and supports resampling, spatial shifts, and cross-benchmark transfer. Across more than 1,000 runs, we profile...

    arxiv.org/abs/2610.08680 · PDF

  10. 10

    Variance-Optimal Off-Policy Evaluation with Conjunct Effect Modeling

    Nicolò Felicioni, Michael Benigni, Maurizio Ferrari Dacrema, Paolo Cremonesi

    cs.LG

    Off-policy evaluation (OPE) for contextual bandit policies becomes challenging when action-level importance weighting incurs excessive variance. Doubly robust (DR) estimation remains unbiased under common support but retains these high-variance action-level weights. A prior estimator, Off-policy evaluation with Conjunct Effect Model (OffCEM), replaces them with more stable cluster-level weights, at the cost of relying on local correctness of...

    arxiv.org/abs/2610.08677 · PDF

  11. 11

    Principled Under Pressure: Post-Training Decides Whether LLMs Act on Their Own Moral Judgment

    Orion Reblitz-Richardson

    cs.LG · cs.AI · cs.CL

    Language models increasingly act as agents. An agent that says an action is wrong and then takes it anyway is a different failure from one that does not know better, and evaluations of stated values cannot see it. We build a pre-registered panel of 248 scenarios across five kinds of pressure. Each scenario is posed twice to the same model, once as the agent choosing what to do and once in the third person asking which option is right, so the...

    arxiv.org/abs/2610.08670 · PDF

  12. 12

    MemFLoRA: Memory-Floor LoRA for CNN Adaptation at the Edge

    Mehmet Emre Akbulut, Johannes Geier, Ulf Schlichtmann

    cs.LG · cs.AI

    On-device learning is necessary when the model encounters user-,sensor-, or environment-specific shifts after deployment. Although parameter-efficient fine-tuning (PEFT) methods, particularly Low-Rank Adaptation (LoRA) variants, enable efficient adaptation at the edge, the limiting resource for Convolutional Neural Network (CNN) adaptation is often not the number of trainable parameters but the activation state that must be retained until the...

    arxiv.org/abs/2610.08669 · PDF

  13. 13

    Early Memory Selection for Balanced Adam

    Alberto Fernández-Hernández, Cristian Pérez-Corral, Jose I. Mestre, Manuel F. Dolz, Enrique S. Quintana-Ortí

    cs.LG · cs.AI · stat.ML

    We propose a method for choosing the shared memory parameter $β_1=β_2=β$ in Adam from a short pilot training. The selected $β$ remains fixed during the subsequent full training. A local model of Adam's normalized direction balances sampling variability against the delay introduced by averaging past gradients. This balance gives a cubic memory rule, whose two coefficients are estimated from gradient probes at a few pilot checkpoints. The...

    arxiv.org/abs/2610.08624 · PDF

  14. 14

    Multi-Label Perceptual Bug Detection in Video Games using Deep Learning on Gameplay Footage

    Nahian Rifaat, Felix Morosov, Loutfouz Zaman

    cs.LG

    Traditional approaches for automated bug detection in video games, such as manual testing, can be beneficial for the improvement of quality assurance, but they can be expensive and time-consuming. The scarce number of tools available to detect multiple perceptual bugs in the same video frame introduces detection challenges for automated bug detection tools in real-world scenarios. We propose a deep learning model for multi-label perceptual...

    arxiv.org/abs/2610.08593 · PDF

  15. 15

    CNet: A Complex-Valued Deep Learning Framework with Wirtinger Autodifferentiation and FFT--Hadamard Convolution

    Marcel Crasmaru

    cs.LG · cs.MS · math.OC

    CNet is a C++/CUDA framework for building and training deep complex-valued neural networks (CVNNs) and, more generally, for optimizing complex-valued functions by gradient descent with Wirtinger (CR-calculus) derivatives. It takes a physics-native stance: a network is a cascade of complex -- and often unitary (the DFT) -- operations acting on an amplitude vector, and classification is a Born-rule measurement $p_k = |z_k|^2 / \|z\|^2$ rather...

    arxiv.org/abs/2610.08592 · PDF

  16. 16

    Random Feature Gaussian Process Attention: Linear-Time Probabilistic Attention with Calibrated Uncertainty

    Amir Mohammad Mahfoozi, Zi Yang, Ying Li, Michael Minyi Zhang

    cs.LG

    Transformers provide a state-of-the-art modeling framework, yet poor calibration limits their reliability in safety-critical applications. A promising direction addresses this issue by interpreting attention as a Gaussian process (GP) posterior, which enables principled uncertainty calibration but incurs cubic complexity in sequence length due to the inversion of the kernel; although decoupled GP variants reduced the cost to quadratic, the...

    arxiv.org/abs/2610.08578 · PDF

  17. 17

    How Learning Governs Unlearning across the Memorization-Generalization Spectrum

    Hwiyeong Lee, Hyelim Lim, Ingyu Bang, Hoki Kim, Taeuk Kim

    cs.LG · cs.AI

    While unlearning seeks to negate undesired capabilities acquired through learning, little research has examined how the way models learn shapes their subsequent unlearning. In this paper, we investigate this connection from the perspectives of memorization and generalization, the two most representative yet competing strategies that models employ during training. We first classify memorization- and generalization-heavy models using grokking...

    arxiv.org/abs/2610.08577 · PDF

  18. 18

    Less Is More: A Leakage-Controlled Study of Dermoscopic Preprocessing for Joint Skin Lesion Classification and Segmentation with YOLO26

    Truong Viet Vu, Nguyen Chi Hai, Nguyen Phuc Nguyen, Ngo Hoang Tu, Vo Nguyen Quoc Bao, Nguyen Thai Anh

    cs.LG · cs.CV

    Handcrafted preprocessing is widely employed in automated dermoscopic analysis to suppress imaging artifacts and enhance lesion visibility. Nevertheless, its actual contribution to modern real-time models remains unclear, particularly when evaluation protocols do not adequately control correlations among images of the same lesion. This study presents a leakage-controlled, lesion-disjoint evaluation of dermoscopic preprocessing and...

    arxiv.org/abs/2610.08570 · PDF

  19. 19

    Singular Value Decomposition: A Geometric Rediscovery, Where Proofs Become Algorithms

    Paul Agron

    cs.LG · math.HO

    This article is a geometric rediscovery of the singular value decomposition, with a further claim: the construction it builds is the machinery behind much of machine learning. The same argument that answers an idle question about ellipses is the algorithm behind principal component analysis, kernel methods, and PageRank, and it is not only the results that transfer but the proofs themselves, run as procedures. The usual introduction states $A...

    arxiv.org/abs/2610.08565 · PDF

  20. 20

    Valid for Free: Homophily-Gated Conformal Prediction for Training-Free Node Classification with Tabular Foundation Models

    Nguyen Duy Long, Phung Minh Hien, Nguyen Trong Viet, Nguyen Thai Anh

    cs.LG

    Tabular foundation models (TFMs) can classify the nodes of a graph without training on it, by reading node and neighborhood features as table rows next to labeled context rows. Work in this line reports predictive performance, not conformal coverage or prediction-set size. To our knowledge, we give the first reliability study of the setting, with TabICL as the TFM and half of each graph as labeled context. As for any predictor fixed before...

    arxiv.org/abs/2610.08564 · PDF

  21. 21

    Reinforcement Learning for Hierarchical Reasoning Rewards: Minimax-Optimal Rates with Transformers

    Naoki Nishikawa, Taiji Suzuki

    cs.LG · stat.ML

    Reinforcement learning (RL) has become a standard tool for post-training language models on reasoning tasks, where the policy is updated by reward feedback while exploring the space of responses. Despite its empirical success, theoretical understanding of RL post-training remains limited, in particular of why on-policy exploration combined with a neural reward model is effective. In this paper, we address this question by modeling the reward...

    arxiv.org/abs/2610.08561 · PDF

  22. 22

    DeltaTTT: Layerwise Optimization for Nonlinear Recurrent Memory

    Yining Li, Dongchen Han, Jie Fu, Gao Huang

    cs.LG · cs.AI · cs.CL

    Sequential test-time training adapts a memory network through successive updates, each computing an inner-loop gradient based on the network's previous state. Intuitively, this state dependence should allow each update to account for what the memory has already learned and better incorporate new information. However, we find that this expected advantage does not consistently materialize in nonlinear memories: a fixed-base parallel TTT...

    arxiv.org/abs/2610.08553 · PDF

  23. 23

    From Shared Demand Patterns to Local Uncertainty: Probabilistic Load Forecasting by Mixing Compact Adaptations

    Haoran Li, Zhe Cheng, Yang Weng

    cs.LG · cs.AI

    Probabilistic load forecasting has been widely studied for power-system operation and planning, but customer- and transformer-level forecasting introduces a distinct scalability challenge. At these levels, load uncertainty is strongly affected by customer behavior, weather, and mixed load composition, making it difficult for a single shared model to capture heterogeneous patterns. Using separate probabilistic models can improve local...

    arxiv.org/abs/2610.08538 · PDF

  24. 24

    FlowCF: Sparse Counterfactual Explanations for Mixed-Type Tabular Data using Flow Matching

    Emmanouil Panagiotou, Eirini Ntoutsi

    cs.LG · cs.AI

    In the field of Explainable AI (XAI), counterfactual (CF) explanations interpret a model's decision by suggesting the changes to the input that would lead to a more favourable outcome. To be useful in practice, such an explanation should change few features and change them as little as possible, properties known as sparsity and proximity. We observe that existing methods remain limited in this respect, especially for numerical features,...

    arxiv.org/abs/2610.08537 · PDF

  25. 25

    How Bregman Divergences Shape Shampoo

    Bing Liu, Wenjie Zhou, Chengcheng Zhao, Hongtao Zhang, Boao Kong, Felix Dangel, Wu Lin

    cs.LG

    Understanding the principles behind Shampoo has recently guided the development of more effective neural network optimizers. These methods learn a preconditioner by optimizing the Frobenius or Kullback-Leibler (KL) divergence against the gradient second moment. In this work, we investigate how the choice of divergence shapes preconditioning, which remains unclear and blocks further improvements. To do so, we develop a unified Bregman...

    arxiv.org/abs/2610.08534 · PDF

  26. 26

    PHBA: Prefix-State Hybrid Block Attention

    Ruijie Li, Jiaxi Hu, Shiyu Wang, Yuxuan Liang

    cs.LG

    Hybrid architectures combining linear sequence models with softmax attention provide an effective balance between efficient long-context modeling and precise token retrieval. Existing designs such as Native Hybrid Attention (NHA) combine compressed long-term states with sliding-window attention, but their exact attention is restricted to a fixed local window. In this work, we introduce Prefix-State Hybrid Block Attention (PHBA), which...

    arxiv.org/abs/2610.08527 · PDF

  27. 27

    MetaLearnNCA: Few-Shot Offline Meta-Learning via Interacting Neural Cellular Automata

    Etienne Guichard, Stefano Nichele

    cs.LG · cs.AI

    Few-shot meta-learning traditionally formulates task adaptation either as analytical gradient descent through unrolled computational graphs or as metric-based distance comparisons over flattened 1D fea- ture vectors, which either incur costly test-time backpropagation or discard native 2D spatial geometry. In this work, we propose METALEARNNCA, a decentralized framework that achieves few-shot adapta- tion through the dynamical interaction of...

    arxiv.org/abs/2610.08479 · PDF

  28. 28

    Learning PDE solution operators with variable initial conditions via Latent Dynamics Networks

    Stefano Maria Pizzamiglio, Stefano Pagani, Francesco Regazzoni

    cs.LG · math.NA

    In many-query scenarios, data-driven surrogate models provide an efficient alternative to high-fidelity solvers for simulating physical systems governed by Partial Differential Equations (PDEs). In this context, the Latent Dynamics Network (LDNet) has recently demonstrated remarkable performance in predicting the response of spatio-temporal systems, combining Neural Ordinary Differential Equations with nonlinear dimensionality reduction....

    arxiv.org/abs/2610.08475 · PDF

  29. 29

    Symmetry-Aware Feature Learning: A Polynomial Separation for Multi-Index Models

    Jivan Waber, Vanessa Piccolo, Yatin Dandi, Florent Krzakala

    cs.LG · math.PR · stat.ML

    We establish a polynomial sample complexity separation between symmetry-aware and symmetry-agnostic feature learning. We study growing-rank multi-index models with high-dimensional Gaussian covariates in $\mathbb{R}^d$ and $r=Θ(d^δ)$ teacher directions forming a cyclic symmetry orbit, where $0<δ<1/2$. We compare three ways of exploiting this structure: architectural weight sharing, data augmentation over the full symmetry group, and learning...

    arxiv.org/abs/2610.08420 · PDF

  30. 30

    SSR: Sparse Segment Reduction for Ternary GEMM Acceleration

    Adeline Pittet, Shien Zhu, Valérie Verdan, Gustavo Alonso

    cs.LG

    Large Language Models (LLMs) require substantial computational resources, limiting their deployment on resource-constrained hardware. Ternary LLMs mitigate these demands through weight quantization via ternary values, achieving significant compression often with 50-90% sparsity. However, existing approaches have limitations: methods optimized for ternary weights, such as BitNet, redundant segment reduction (RSR), and its improved version...

    arxiv.org/abs/2610.08403 · PDF

  31. 31

    VETTA: Coordinating Turn- and Token-Level Credit Assignment for Multi-Turn LLM Agents

    Jiaju Chen, Min Yang, Jinghua Piao, Xiaochong Lan, Xu Xia, Xiangnan He, Yong Li

    cs.LG

    Multi-turn LLM agents often receive sparse task feedback across several interactions, while generating each response token by token. This creates two related credit-assignment questions: which responses helped achieve the outcome, and which generation decisions mattered within each response? Existing methods typically focus on only one level: turn-level methods evaluate complete responses but do not distinguish the decisions within them;...

    arxiv.org/abs/2610.08402 · PDF

  32. 32

    Atom-JEPA: Joint-Embedding Predictive Architecture for 3D Atomistic Systems

    Kasper Helverskov Petersen, Rasmus Hannibal Tirsgaard, François R J Cornet, Mikkel Jordahn, Mikkel N. Schmidt

    cs.LG · cs.AI · physics.comp-ph

    Large-scale self-supervised pretraining has reshaped modern machine learning, substantially advancing the ability of language and vision models to generalize across downstream tasks. While deep learning has driven considerable progress in modeling atomistic systems in recent years, self-supervised pretraining in this domain has not yet achieved comparable downstream generalization. To address this, we introduce Atom-JEPA, a self-supervised...

    arxiv.org/abs/2610.08400 · PDF

  33. 33

    Decision-Focused Learning in MDPs: An Occupancy Measure Approach

    Zihao Zhao, Ashwath K. Karunakaram, Ali Eshragh, Yuexing Li, Kai Wang

    cs.LG

    In this work, we consider decision-focused learning (DFL) for a Markov decision process (MDP), where existing methods differentiate through the KKT conditions of the Bellman equation and require solving a linear system over all state-action pairs, limiting its scalability. We address this by reformulating the MDP as an occupancy measure-based linear program (LP), whose feasible region is induced by predicted dynamics, and we derive a...

    arxiv.org/abs/2610.08384 · PDF

  34. 34

    Accelerating the Development of PLGA In Situ Forming Depots Through AI-Driven Multi-Objective Optimization

    Pauric Bannigan, Siddarth Chandrasekaran, Brigitte A. G. Lamers, Inge Hermsen, Gary Tom, Riley J. Hickman, Bahar...

    cs.LG · cs.AI

    Developing long-acting injectable formulations requires the simultaneous optimization of drug loading, release kinetics, viscosity, injectability, stability and other objectives. To navigate this multidimensional space, Corbion and Intrepid combined Corbion's diverse PURASORB bioresorbable polymer library with Intrepid Labs' proprietary AI algorithm (ANDROMEDA 1) to develop in situ forming depots for a therapeutic peptide. Over approximately...

    arxiv.org/abs/2610.08368 · PDF

  35. 35

    Evolutionary One-Step Generators: Fast and Diverse Sampling for Discrete Design

    Marcus Vukojevic, Erik Nielsen, Veronica Lachi, Andrea Passerini, Giovanni Iacca

    cs.LG · cs.NE

    Several discrete design tasks, such as molecular discovery, require diverse collections of useful candidates at low computational cost. High validity alone does not guarantee a useful candidate library: repeatedly generating the same valid structures leaves few distinct alternatives. Training for both feasibility and diversity is challenging because many relevant criteria can only be evaluated after hard decoding. To address this challenge,...

    arxiv.org/abs/2610.08367 · PDF

  36. 36

    Sensor Geometry as a Flow-Matching Prior for Multi-Channel Brain Signals

    Jaedong Hwang

    cs.LG · cs.AI

    Flow-matching models start from an isotropic Gaussian source, the standard choice when the correlation structure of the data is unknown in advance. For multi-channel brain recordings, however, part of this structure is known in advance. Electrodes sit at fixed positions on the head, and volume conduction through the skull and scalp makes nearby electrodes co-vary in a way that is shared across subjects. Existing EEG generative models...

    arxiv.org/abs/2610.08355 · PDF

  37. 37

    Uncertainty Quantification Is Indispensable for Reliable Connectome-Based Graph Learning: A Narrative Review and Case Study

    Mansooreh Pakravan

    cs.LG

    While graph neural networks (GNNs) have shown substantial promise in connectome-based diagnostic classification, deterministic models inevitably suppress pipeline-induced noise and model ambiguities, yielding overconfident predictions. Although uncertainty quantification (UQ) is widely adopted in voxel-level segmentation, its role in connectomic graph learning remains largely unaddressed. This paper presents a comprehensive narrative review...

    arxiv.org/abs/2610.08353 · PDF

  38. 38

    Structure-Aware Graph Abstention for Reliable Selective Forecasting

    Jianxiang Xie, Belal Alsinglawi

    cs.LG

    Selective forecasting abstains on high-risk test windows under a retained-coverage budget. Existing gates such as TEM (Brusokas et al., 2025) score each forecast as a whole; for multivariate outputs, trajectories can look plausible while violating dependencies among variables. We treat instance-level plausibility and relational consistency as distinct reliability axes and operationalize the latter via a learned sparse graph and a...

    arxiv.org/abs/2610.08322 · PDF

  39. 39

    OxiGen: Oxidation-State-Aware Crystal Generation

    Dylan John, Kim E. Jelfs, Alex M. Ganose, Eleonora Giunchiglia

    cs.LG · cond-mat.mtrl-sci

    Generative models have the potential to accelerate inorganic materials discovery by enabling inverse design, but generating experimentally realisable crystals remains challenging. Oxidation states are widely used to assess the compositional validity of crystals and guide inorganic materials discovery. While existing generative models for crystals can generate materials with charge-neutral oxidation-state assignments, they poorly reproduce the...

    arxiv.org/abs/2610.08296 · PDF

  40. 40

    Scalable extraction and visualization of multi-attribute logical and functional dependencies in tabular data

    Chaithra Umesh, Arvind Lomrore, Neethu D, Kristian Seegel-Schultz, Saptarshi Bej, Olaf Wolkenhauer

    cs.LG

    Understanding the structural relationships among attributes in tabular data is fundamental to machine learning and pattern recognition. While functional dependency (FD) discovery has been extensively studied, scalable discovery of logical dependencies (LDs), particularly as the number of attributes and dependency order increase, remains underexplored. These dependencies capture non-deterministic, condition-specific relationships among...

    arxiv.org/abs/2610.08287 · PDF

  41. 41

    Performative Prediction with Selective Labels

    Giovani Valdrighi, Isabel Valera, Marcos Medeiros Raimundo

    cs.LG

    Many social applications of machine learning exhibit performative effects: population behavior changes in response to deployed models. Performative prediction studies this interaction through a distribution map that relates each model to the population distribution it induces. One of the main results in this framework showed that repeated risk minimization (RRM), which updates models by retraining on the most recent data, can converge to a...

    arxiv.org/abs/2610.08272 · PDF

  42. 42

    Reinforcement Learning with Segment Reward Feedback under Linear Function Approximation

    Fengxu Liu, Siwei Wang, Gal Dalal, Shie Mannor, Yihan Du

    cs.LG

    Classical reinforcement learning (RL) assumes that a reward is observed for every visited state-action pair. However, in real-world applications such as autonomous driving, such fine-grained feedback can be costly or difficult to collect, whereas trajectory-level feedback may be too sparse for efficient learning. To provide a general feedback model bridging these two extremes and handle large state spaces, we study RL with segment reward...

    arxiv.org/abs/2610.08271 · PDF

  43. 43

    Finding the Heads and the Neurons Responsible for Network Information Retrieval in Language Models

    Abdul Kadir, Md Mohasin Hossain, Daniel Sonntag

    cs.LG

    We ask whether specific attention heads, and more finely specific neurons inside those heads, are responsible for recognizing that a language model's context contains network infrastructure information (a hostname paired with its IP address), and whether that responsibility can be validated causally rather than by correlation alone. At the head level the answer is yes, across five models spanning three architecture families: in every model, a...

    arxiv.org/abs/2610.08200 · PDF

  44. 44

    On the Intrinsic Limited Robustness of Latent-Based Watermarking

    Cheng-Han Yeh, Kuan-chun Yu, Cheng-Chang Tsai, Chun-Shien Lu

    cs.LG · cs.CR

    Existing latent-based watermarking methods for diffusion models have overestimated their robustness to image distortions, including geometric transformations such as rotation, scaling, and translation (RST). Moreover, this paradigm of watermarking approaches may suffer from inherent limitations arising from the domain in which the watermark is embedded. In this paper, we provide the first theoretical analysis explaining why these methods lack...

    arxiv.org/abs/2610.08178 · PDF

  45. 45

    Align, Then Correct: Training-Free Two-Stage Low-Rank Compensation for Extremely Quantized Large Language Models

    Seobin Song, Geonho Lee, Janghwan Lee, Jungwook Choi

    cs.LG · cs.CL

    Low-rank quantization error compensation (LQEC) recovers the accuracy lost under aggressive weight quantization by attaching a closed-form rank-$r$ adapter beside each frozen quantized weight, without any training. We show that existing compensators are limited by two shared simplifications. They calibrate symmetrically, evaluating the full-precision and compensated weights on the same activation, which yields a compensation target that is...

    arxiv.org/abs/2610.08164 · PDF

  46. 46

    Symphony for Text Generation: Benchmarking Clinical Note Generation

    Daniel Varab, Victor Petrén Bach Hansen, Asbjørn W. Helge, Kevin Pelgrims, Mathias Baltzersen, Adrian Young-San...

    cs.LG · cs.AI · cs.CL

    Ambient documentation systems are rapidly gaining adoption, yet their impact on clinical note quality remains poorly characterized. We introduce MedConv, a multilingual dataset of 300 clinical encounters in English, Danish, and German, and use it alongside the Ambient Clinical Intelligence benchmark (ACI-BENCH) to compare Corti, a clinical AI platform, with two leading, accessible ambient scribe software applications built on general-purpose...

    arxiv.org/abs/2610.08161 · PDF

  47. 47

    Mu-DisCoCat: A Variational Pipeline for Compositional Generalization on Quantum Processors

    Mina Abbaszadeh, Matilda Karabina Moore, Raem Haq, Martha Lewis, Mehrnoosh Sadrzadeh

    cs.LG · cs.CV

    Achieving compositional concept generalization (CoCoGen), the ability to understand novel situations by recombining learned primitives, remains a fundamental challenge in artificial intelligence. Compositional semantic models such as Compositional Distributional Semantics (DisCoCat) offer solutions by generalising vectors to tensors, but suffer from scaling bottlenecks when learning the tensors. Mapping DisCoCat onto Variational Quantum...

    arxiv.org/abs/2610.08131 · PDF

  48. 48

    Do LLMs Act on What They Know? From Partner Representations to Cooperative Actions

    Yuhwan Jeong, Jinnyeong Yang, Kuk-Jin Yoon

    cs.LG

    Cooperation with unfamiliar partners requires adapting to communication conventions that are not known in advance. We study this problem in a controlled Hanabi-derived environment with scripted hint generation, LLM-controlled receiving decisions, and frozen model weights. Across eight LLMs, linear probes recover intent conventions substantially more accurately than target conventions, yet receiving choices do not consistently agree with the...

    arxiv.org/abs/2610.08129 · PDF

  49. 49

    Attenuated in-context identification in time-series foundation models: diagnosis under counterfactual inputs and repair by synthetic forced-system fine-tuning

    Hong-In Won

    cs.LG · eess.SY

    Covariate-aware time-series foundation models (TSFMs) promise training-free what-if answers for instrumented plants: the change in output that a different future input would cause. We test this on forced engineering systems with exact counterfactuals, comparing Chronos-2, TimesFM-2.5 and TabPFN-TS with classical system identification fitted to the same context. Through their default covariate interfaces, TimesFM-2.5 and TabPFN-TS are...

    arxiv.org/abs/2610.08118 · PDF

  50. 50

    Enhancing Diffusion Language Models with Autoregressive Post-Training Weights

    Yiming Qin, Ke Wang, Amel Abdelraheem, Adam Hazimeh, Pascal Frossard

    cs.LG

    Diffusion language models (dLLMs) have emerged as a promising alternative to autoregressive (AR) language models, offering flexible token-update orders and parallel decoding. Recent dLLMs are often initialized from pretrained AR models before diffusion conversion in order to inherit their learned representations. After the conversion, however, they typically ignore the extensive post-training ecosystem of their AR ancestors. In this work, we...

    arxiv.org/abs/2610.08108 · PDF

  51. 51

    Optimization Encoders: Rethinking Second-Order Meta-Learning for Neural Fields

    Rudolf L. M. van Herten, Soufiane Ben Haddou, Rachit Saluja, Johannes C. Paetzold

    cs.LG · cs.CV

    Conditional neural fields represent signals continuously, but their effectiveness depends on how the conditional latent representations are inferred from observed data. In meta-learning, this encoding occurs through gradient updates induced by the decoder, tying representation learning directly to decoder design. We formalize this connection by interpreting latent optimization as an optimization encoder, unifying the roles of second-order...

    arxiv.org/abs/2610.08075 · PDF

  52. 52

    Detecting a Shift Is Not Enough: Exact Minimax Limits of Linear Representation Repair

    Anuar Aimoldin, Yankai Chen, Ayana Mussabayeva, Xue Liu

    cs.LG · math.ST · stat.ML

    A mean shift between two data sources can be easy to detect but hard to remove without substantially changing their representations. We cast its removal as a statistical decision problem: from noisy differences between paired calibration measurements in $\mathbb{R}^d$, learn one linear map, applied to both sources under a hard distortion budget, that leaves as little of the shift as possible on fresh data. We derive the exact finite-sample...

    arxiv.org/abs/2610.08069 · PDF

  53. 53

    A Riemannian Geometry for Low-rank Adaptation

    Shoichiro Takeda, Shin'ya Yamaguchi, Satoshi Suzuki, Yasunori Akagi

    cs.LG

    Low-rank adaptation (LoRA) is widely used as a parameter-efficient fine-tuning technique for pre-trained deep neural networks, which approximates the weight update via full fine-tuning by a low-rank matrix $BA^\top$. This parameterization leads to the equivalence relation $(B, A) \sim (BG^{-1}, AG^\top)$ for any invertible matrix $G$ because $BA^\top = BG^{-1}(AG^\top)^\top$ and thus both pairs yield the same loss value. This relation induces...

    arxiv.org/abs/2610.08049 · PDF

  54. 54

    SepsisLens: Structure-Preserving Sequence Modelling for Decomposable Early Sepsis Warning

    Yikun Ou, Wei Li

    cs.LG

    Early sepsis warning from ICU records can be cast as a structure-preserving prediction problem. A model needs to detect deterioration from irregular measurements while keeping each alert connected to the physiological signals that support it. Many temporal models fuse clinical variables into a patient-level representation, supporting scalar risk prediction but weakening the structure needed for clinical decomposition. We present SepsisLens,...

    arxiv.org/abs/2610.08046 · PDF

  55. 55

    Spectra: Exact Component Transport for Test-Time Prior Adaptation in Simulation-Based Inference

    Xin Zhao, Nico Scherf, Robert Trampel, Kerrin J. Pine, Nikolaus Weiskopf

    cs.LG · stat.ML

    Simulation-based inference (SBI) has become a powerful approach to Bayesian inference in complex scientific models whose likelihoods are difficult or impossible to evaluate. Amortized SBI learns reusable inference models from simulated data, enabling rapid posterior inference for new observations, and modern generative models have made these models increasingly expressive. However, this reuse is limited to the prior distribution chosen during...

    arxiv.org/abs/2610.08021 · PDF

  56. 56

    Can phenotypic activity be predicted without experimental readouts?

    Télio Cropsal, Rocío Mercado

    cs.LG

    Molecular encoders contrastively pretrained on paired molecule-morphology data, such as CLOOME and CellCLIP, have been proposed as cheap surrogates for phenotypic prediction, avoiding the need to run a Cell Painting assay. We evaluate this idea for these molecular encoders under a protocol designed to control for two confounds that can inflate apparent performance: leakage across an encoder's own pretraining boundary, and the correlation...

    arxiv.org/abs/2610.07997 · PDF

  57. 57

    TICDA: Tabular In-Context Data Attribution

    Yacine Benihaddadene, Milan Bhan, Eliot Dugelay, Mohammed Jawhar, Benjamin Wong, Nicolas Chesneau, Duong Nguyen

    cs.LG · cs.AI

    Tabular foundation models (TFMs) achieve strong predictive performance by conditioning on labeled demonstrations provided in context, without any parameter update. Yet how individual demonstrations shape a given prediction remains poorly understood. This gap matters in practice: the context is often assembled from whatever labeled data is available, potentially leading to the inclusion of mislabeled, redundant, or low-quality examples that...

    arxiv.org/abs/2610.07996 · PDF

  58. 58

    A Broader Look at Model Merging: Rethinking Implicit Regularization Induced by Task Arithmetic

    Sin-Han Yang, Shih-Cheng Huang, Chieh-Yen Lin, Yun-Nung Chen, Shao-Hua Sun, Hung-yi Lee

    cs.LG · cs.CL · cs.CV

    Model merging aims to build a multi-task model cheaply by combining the weights of individual task-specific models. To perform well across multiple tasks, most existing merging methods use an additional dataset to find the coefficients for the best linear combination of task-specific weight updates. However, we identify an implicit regularization in this standard practice: searching over coefficients restricts the candidate models to a...

    arxiv.org/abs/2610.07990 · PDF

  59. 59

    Do Higher-Order Models Win for Higher-Order Reasons? Rethinking Performance Gains in Hypergraph Learning

    Fanchen Bu, Fan Li, Geon Lee, Sunwoo Kim, Xiaoyang Wang, Renaud Lambiotte, Kijung Shin

    cs.LG · cs.SI

    Higher-order models (e.g., hypergraph neural networks) often outperform lower-order baselines on hypergraph learning benchmarks, and their advantages are commonly attributed to their ability to exploit higher-order information. However, better performance alone does not establish this explanation. We therefore ask: Do higher-order models win for higher-order reasons? To investigate this question, we introduce a controlled...

    arxiv.org/abs/2610.07981 · PDF

  60. 60

    Learning a Ranking from Human Feedback in Log-Concave Random Utility Models

    Diego Alovisetti, Marco Mussi, Alberto Maria Metelli

    cs.LG

    We study the problem of recovering the ranking of a fixed set of items according to their unknown numerical utilities. At each interaction with the environment, a learner presents the item set to a human and receives comparative feedback of two types. Under full-ranking feedback, each interaction reveals a noisy ranking of all items, whereas under winner-only feedback, it reveals only the item ranked first. In both settings, we model human...

    arxiv.org/abs/2610.07973 · 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.