cs.LG · 2026-09-22 · No. 121

Machine Learning, 2026-09-22.

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

01 — The papers

66 entries
  1. 01

    Critical-State RL: Diagnosing Trainable States for Multi-Turn Tool Use

    Zixiang Chen, Wenting Zhao, Zhepeng Cen, Akshara Prabhakar, Jielin Qiu, Jianguo Zhang, Zhiwei Liu, Tulika Manoj...

    cs.LG · cs.CL

    Multi-turn tool-use failures can hinge on a single model call, yet reward variation alone does not reveal which call would benefit from training. When rewards depend on later interactions, their variation can reflect downstream randomness rather than differences between the current actions. We introduce Critical-State RL to identify trainable states in multi-turn interactions. Given task-defined candidate calls and local rewards, the method...

    arxiv.org/abs/2609.24985 · PDF

  2. 02

    LoRA-generating hypernetworks for efficient on-device LLM generative personalization

    Sean Augenstein, Li Ding, Jihwan Lee, Keith Rush, Andrey Zhmoginov

    cs.LG

    On-device large language models (`LLMs'), e.g. running on mobile phones, are ripe for improvement via personalization. The limited compute resources of mobile devices impose limits on model scale and thus model quality, making any realizable quality gains highly impactful. At the same time, their personal nature (i.e., the close coupling to a particular user) means that a given on-device LLM tends to be used in similar, predictable patterns...

    arxiv.org/abs/2609.24979 · PDF

  3. 03

    RRSI: Regularized Recursive Self-Improvement of Agent Harnesses

    Peng Xia, Rujun Han, Zifeng Wang, Yanfei Chen, Yufan Zhang, Yoonho Lee, Chengsong Huang, Han Yu, Zhongying CuiZhu,...

    cs.LG · cs.AI · cs.CL

    An LLM agent's capability is largely magnified by its harness, namely the prompts, control flow, tooling, memory, and context management surrounding the frozen backbone model. Recent methods increasingly automate this process by iteratively proposing and selecting component-wise edits of an agent harness, practically establishing a form of recursive self-improvement (RSI) at the agent-system level. However, such recursive evolution may...

    arxiv.org/abs/2609.24972 · PDF

  4. 04

    Rare Event Estimation via Iterative Unalignment

    Hanming Yang, Daksh Mittal, Jing Dong, Hongseok Namkoong

    cs.LG · cs.AI

    As agents are deployed with increased autonomy, even extremely rare events along their stochastic output trajectories can occur and prove catastrophic. Safe deployment therefore does not depend on whether these events can occur, but on how often they might. We study the problem of estimating the probability of rare events that arise from stochastic variation in the agent's own actions. Estimating this type of risk requires searching over the...

    arxiv.org/abs/2609.24969 · PDF

  5. 05

    Learning Physics from an Imperfect Ancestor

    S. Mohammad Mousavi, Teeratorn Kadeethum, Nikolaos Bouklas, Somdatta Goswami

    cs.LG

    Neural operators evaluate parametric partial differential equations cheaply but degrade sharply outside their training distribution. Physics-informed neural networks avoid dependence on labeled data, yet their optimization can be basin-fragile: when the governing residual admits multiple solutions, a PINN trained from scratch may converge to a physically incorrect state despite achieving a small residual. We show that these failure modes can...

    arxiv.org/abs/2609.24947 · PDF

  6. 06

    Exactness at Inference: A Representational Criterion for Out-of-Distribution Generalization

    Filipe Marinho Rocha, Inês Dutra, Vítor Santos Costa, Luís Paulo Reis

    cs.LG · cs.AI · cs.LO

    A model generalizes outside its training distribution only when it computes a representation structurally equivalent to the generating mechanism, not an approximation fitted to it. Such equivalence is necessary for exactness in and out of distribution, and extrapolation is governed by this exactness at inference, whatever its realization. Tensor Logic shows this: a zero-temperature contraction is equivalent to discrete logic, deducing in...

    arxiv.org/abs/2609.24942 · PDF

  7. 07

    Learning Prognostic Variables for AI Convective Parameterizations via Symbolic Distillation

    Jurij Schönfeld, Tom Beucler, Julien Savre, Steven Sherwood, Veronika Eyring

    cs.LG · physics.ao-ph

    Hybrid AI-physics climate modeling aims to improve coarse (~100km-resolution) Earth system models by learning to parameterize subgrid processes from high-fidelity data. However, this so far mostly involves local-in-time, diagnostic parameterizations, in which the subgrid state depends only on the current coarse state with no memory of previous states, which is unrealistic for processes such as convection that have intrinsic persistence. To...

    arxiv.org/abs/2609.24882 · PDF

  8. 08

    When Tomorrow Becomes Today: Self-Evolving Policies for Agentic Time-Series Forecasting

    Yifan Hu, Xilin Dai, Zhiyuan Qu, Yiding Liu, Zewei Dong, Jiang-ming Yang, Qiang Xu

    cs.LG · cs.AI

    Agentic time series forecasting concerns systems whose underlying mechanisms evolve, making the relative effectiveness of numerical models, reasoning strategies, and intervention rules inherently time-varying. Consequently, a time series agent must adapt the forecasts it produces and the orchestration policy that determines which components to trust and how to coordinate them. The deployment process naturally provides supervision for this...

    arxiv.org/abs/2609.24862 · PDF

  9. 09

    G-NAC: Graph Neural Automata Clustering via Emergent Domain Formation

    Keith Miller, Tristan Crawford

    cs.LG · eess.SP

    We introduce Graph Neural Automata Clustering (G-NAC), an unsupervised clustering method in which observations interact as cells on a fixed neighborhood graph. A shared recurrent graph-neural cellular rule evolves latent domain states through local interactions, which are converted into a rank-based spectral affinity for partitioning. Across 73 clustering tasks from 57 benchmark datasets, G-NAC achieved a mean adjusted Rand index (ARI) of...

    arxiv.org/abs/2609.24823 · PDF

  10. 10

    Complex KDA: Understanding and Enhancing the Expressivity of Kimi Delta Attention

    Julien Siems, Riccardo Grazzi, Korbinian Pöppel, Jaisidh Singh, Arber Zela, Timur Carstensen, Jenia Jitsev, Frank...

    cs.LG

    Linear RNNs based on the delta-rule enable efficient sequence modeling, but their linear updates with a low-rank correction constrain their expressivity. Prior work has shown that composing two delta-rule transitions in a single recurrent update can model a 2D rotation, but this increases the rank and the cost of the updates compared to a single transition. We show that Kimi Delta Attention (KDA) can realize 2D rotations by combining a single...

    arxiv.org/abs/2609.24797 · PDF

  11. 11

    Inference of Unknown Dynamical Components Using Next Generation Reservoir Computing: From Chaotic Systems to Climate Data

    Jule Budnick, Andrew Keane, Serhiy Yanchuk

    cs.LG · nlin.CD

    We investigate next generation reservoir computing (NGRC) as a data-driven approach for inferring unseen components of dynamical systems. We compare NGRC with traditional reservoir computing (RC) using the Lorenz and Rössler system, where two unknown components are inferred from one given component. For both systems, NGRC achieves accurate results while requiring fewer training data and less computational time than RC. We identified an...

    arxiv.org/abs/2609.24754 · PDF

  12. 12

    Enhancing Transformer Representations of Symbolic ODE Expressions

    Xiyue Fan, Adam Prugel-Bennett, Stuart E. Middleton

    cs.LG · cs.AI

    Existing approaches to solving differential equations, such as symbolic regression, physics informed neural networks, and neural operators, typically focus on numerical approximations or blind symbolic search via fitting to numerical data. Less attention has been paid to learning structured representations of mathematical expressions that preserve commutative properties and could support mathematical reasoning in symbolic forms. Transformer...

    arxiv.org/abs/2609.24746 · PDF

  13. 13

    An Exact Junction-Tree Extended Formulation for Optimal Classification Trees

    Jiancheng TU, WenqiFan

    cs.LG

    We develop an exact linear programming (LP) formulation for bounded-depth classification trees with binary features, using a junction-tree representation. The formulation is integral and supports recursive subtree optimization. Exact reductions make the model smaller while preserving the optimal value and recovery of an optimal tree. The reduced model supports two solution methods: column generation and message passing. Column generation...

    arxiv.org/abs/2609.24741 · PDF

  14. 14

    A Federated Artificial Intelligence Framework for Optimizing Pancreatic Cancer Treatment - Strategy Update

    Anne-Christin Hauschild, Amirreza Aleyasin, Nils H. Beyer, Lisa Fricke, Jonas Hügel, Maryam Moradpour, Anh-Tien...

    cs.LG

    While a centralized approach involving patient consent to collect and analyze data centrally would theoretically offer the best data quality and predictive performance, it is not always feasible in practice. Federated Learning (FL) architectures have shown to be a very promising approach to use and access distributed disease related resources within the GDPR boundaries. In a previous case report, we described the preconditions at the...

    arxiv.org/abs/2609.24718 · PDF

  15. 15

    Guaranteed Low-Rank Tensor Recovery from Modewise Measurements via Normalized Block-Weighted Riemannian Gradient Descent

    Yushi Zhou, Feng Zhang

    cs.LG

    We consider the recovery of low-multilinear-rank tensors from linear measurements and propose an adaptive block-weighted modewise Riemannian gradient descent method. The method combines memory-efficient modewise measurements with a normalized adaptive weighting strategy for the core and factor components of the Riemannian gradient. The weighting improves convergence without increasing the multilinear-rank bound of the search direction or the...

    arxiv.org/abs/2609.24679 · PDF

  16. 16

    Muon Can Outperform Dedicated Continual Learning Methods

    Sebastian George Sincari, Bogdan Alexandru Gheorghe, Antonio Barbalau

    cs.LG · cs.CL

    Continual learning with Low-Rank Adapters (LoRA) typically mitigates forgetting by penalizing the overlap between a new update and the accumulated past weights, which discourages certain update directions without controlling how an update distributes its energy over the ones that remain. We ask whether that restriction has to be task-aware, or whether a generic one supplied by the optimizer is enough. We train a plain incremental LoRA...

    arxiv.org/abs/2609.24678 · PDF

  17. 17

    Corrective Forcing: Unified Post-Training for Diffusions and Flows in Generative Speech Enhancement

    Qing Yao, Lijian Gao, Qirong Mao

    cs.LG · cs.AI

    Diffusion and flow models, as promising generative paradigms for speech enhancement, face a training--inference mismatch: training uses analytical path states, whereas inference recursively evaluates models on self-generated rollout states along discretized sampling trajectories. This mismatch causes prediction and discretization errors to accumulate. To address it, we introduce Corrective Forcing (CoF), a post-training paradigm that forces...

    arxiv.org/abs/2609.24651 · PDF

  18. 18

    iSDFT: Information-Proximal Self-Distillation for Continual Learning in LLMs

    Ahmed Khaled Khamis, Xiaotong Ji, Hassan Jaber, Rasul Tutunov, Matthieu Zimmer, Jun Wang, Haitham Bou-Ammar

    cs.LG · cs.AI

    On-policy self-distillation fine-tuning (SDFT) learns new skills from demonstrations while reducing forgetting, but it always distils toward the full demonstration-conditioned teacher. This fixes teacher influence at the full-teacher endpoint, providing no control over how much demonstration information should be transferred at each prediction state. We introduce Information-Proximal SDFT (iSDFT), which instead treats the teacher as a...

    arxiv.org/abs/2609.24646 · PDF

  19. 19

    Augmented Hypothesis Testing with Persona-Based LLM Simulations

    Ziyad Benomar, Aymen Al Marjani, Paul Missault, Saab Mansour

    cs.LG · cs.AI · stat.AP

    A/B testing requires large sample sizes, long timelines, and significant costs. When auxiliary predictions of experimental outcomes are available from machine learning models, uncertain prediction quality precludes replacing human experiments entirely, yet these predictions may still contain useful signal. We propose a principled framework for learning-augmented hypothesis testing that leverages predictions of unknown quality to reduce sample...

    arxiv.org/abs/2609.24629 · PDF

  20. 20

    GraphToolbox: A Configurable Python Framework for Graph Neural Network Forecasting

    Eloi Campagne, Yvenn Amara-Ouali, Yannig Goude, Argyris Kalogeratos

    cs.LG · cs.AI

    Electricity forecasting often involves spatially related signals observed over regions, substations, and feeders, and Graph Neural Networks (GNNs) provide a natural way to represent these relations. Building a complete GNN forecasting experiment is nonetheless laborious, because graph construction, model selection, training, aggregation, and interpretation sit in incompatible tools. We present GraphToolbox, an open-source Python framework...

    arxiv.org/abs/2609.24609 · PDF

  21. 21

    Taking a Second Look: Correcting Sea Ice Forecasts with Sparse Observations

    Tianshuo Zhang, Xianglei Xing, Aowen Yang, Jia Gao, Wenzhe Zhai, ShanShan Liu

    cs.LG

    Sea ice forecasts are issued several days ahead, allowing errors to accumulate while new, often sparse sea ice concentration (SIC) observations become available. We find that fixed-propagation errors concentrate near structured, high-gradient ice edges, whereas homogeneous interiors require limited propagation, suggesting that propagation distance should be state dependent. We therefore introduce ECHO (Evidence-guided Correction with...

    arxiv.org/abs/2609.24591 · PDF

  22. 22

    Overlay\_dx - Automating forecasting evaluation

    Long Ngo, Mohammed Amine Chamli, Jonathan Rivalan, Thomas Jaillon

    cs.LG · cs.AI

    Traditional evaluation metrics provides numerical values but often lack comprehensibility, hindering effective differentiation of model performances. Our work addresses this challenge by introducing overlay\_dx, a novel evaluation metric measuring the performance of time series prediction models. Overlay\_dx is a visual metric that represents the percentage of predictions falling within a confidence interval around actual values....

    arxiv.org/abs/2609.24586 · PDF

  23. 23

    Universal Multi-Modal Traceformer: Integrating Heterogeneous Context for Process Event Prediction

    Fabian Spaeh, Jingxing Fang, Shandian Zhe, Bin Shen

    cs.LG

    Event logs arise in a wide range of real-world processes, capturing not only event activities and timestamps but also multi-modal contextual information. Existing event-sequence models, including many temporal point process approaches, primarily model event activities and timestamps while overlooking heterogeneous context, such as numerical measurements, categorical attributes, textual descriptions, and metadata associated with individual...

    arxiv.org/abs/2609.24579 · PDF

  24. 24

    $t_0$: A Time-Series Foundation Model for Forecasting with Context

    Lucas Meyer, Claudio Sole, Huikan Xiang, Nicolas Li, Lucas Franceschino, Arnau Quera-Bofarull, Maarten P. Scholl,...

    cs.LG · cs.AI

    We present $t_0$, a family of open-weights foundation models for forecasting with multivariate context. We release its first two members: $\texttt{t0-alpha}$ and $\texttt{t0-beta}$, respectively 102M and 256M parameters. Both condition their forecasts on target history, past covariates, and known-future covariates, without task-specific retraining. Their transformer layers alternate attention along time and across variates. They produce...

    arxiv.org/abs/2609.24559 · PDF

  25. 25

    On Emergent Capabilities and Model Merging

    Luca Zhou, Emanuele Rodolà

    cs.LG · cs.AI

    Fine-tuned checkpoints and adapters now fill public repositories, and the most common operation applied to these artifacts is model merging: arithmetic on their weights that assembles capabilities cheaply. We ask what this operation does to emergent capabilities: behaviors an artifact carries that were never an explicit training target. Studying two independent testbeds (activation oracles and emergent-misaligned models) across three model...

    arxiv.org/abs/2609.24504 · PDF

  26. 26

    Lifted Bellman Linear Programming for Offline Reinforcement Learning

    Hyukjun Yang, Jongchan Park, Narim Jeong, Donghwan Lee

    cs.LG · cs.AI

    Offline reinforcement learning (RL) typically trains a critic by minimizing a regression loss against bootstrapped value targets stabilized by target networks with exponential moving average (EMA) updates. Multi-step targets incorporate behavior-policy actions and therefore require off-policy correction. We instead impose in-sample Bellman optimality on the critic through inequality constraints. We formulate the Lifted Bellman Linear Program...

    arxiv.org/abs/2609.24489 · PDF

  27. 27

    A Temporal Knowledge Graph for Music Festival Lineup Forecasting

    Julia Gastinger, Thilo Dieing, Christian Meilicke, Heiner Stuckenschmidt

    cs.LG

    Music festival lineups emerge from complex relationships among artists, genres, releases, labels, and past performances, making the prediction of future lineups a natural fit for temporal knowledge graph (TKG) forecasting. In this work, we present a TKG covering 380 festivals over 55 years, comprising more than 90K festival performance quadruples along with information on festivals, artist tours, and artist metadata, and release it as a...

    arxiv.org/abs/2609.24467 · PDF

  28. 28

    RAILS: Retrieval-Augmented Incremental LLM Clustering at Scale

    Armin Oliya, Aleksandra Sawczuk, Radosław Białobrzeski

    cs.LG

    Using a Large Language Model (LLM) as the clusterer at production scale is hard: prompts cannot hold the entire label space, and per-document serial processing does not deliver the throughput real workloads require. We present RAILS, a retrieval-augmented incremental LLM clusterer that turns clustering into a simple loop over a growing label pool and scales through document batching with bounded concurrency. On six public benchmarks RAILS...

    arxiv.org/abs/2609.24464 · PDF

  29. 29

    WPBench: A Comprehensive Benchmark for Wind Power Forecasting

    Yuhan Zhu, Jilin Hu, Xinying Cai, Yingshan Li, Li Ma, Xiangfei Qiu Linsen Li, Kai Zhang, Yao Fu, Weihao Jiang, Bin Yang

    cs.LG · cs.AI

    Accurate, reliable, and deployable wind power forecasting is critical for power system dispatch, renewable energy integration, and electricity market operations. Progress in this field hinges on the ability to empirically and comprehensively benchmark forecasting methods. Yet existing benchmarks fall short of supporting systematic evaluation in four key aspects: 1) limited coverage of wind power scenarios across turbine scale, variable...

    arxiv.org/abs/2609.24444 · PDF

  30. 30

    MUSE: Dependency-Aware Adaptation of a Frozen Vision Backbone for Multivariate Time Series Forecasting

    Xinying Cai, Junkai Lu, Yuhan Zhu, Xiaoyun Yu, Xiangfei Qiu, Jilin Hu

    cs.LG

    Multivariate time-series forecasting is essential to many real-world applications. Recent large vision models (LVMs) offer a promising paradigm by transferring cross-domain visual priors to time-series forecasting. However, existing LVM-based methods face two key challenges: balancing independent visual representation spaces with cross-variable dependency modeling, and adapting vision backbones pretrained on natural images to the distinct...

    arxiv.org/abs/2609.24441 · PDF

  31. 31

    Comparing Latent Concept Formation in State Space Models and Transformers via Sparse Autoencoders

    Rithin Nagaraj, Rupa Laalasa Oruganti, Prerna Subhashchandra Kunder, Ashwini M Joshi

    cs.LG

    The quadratic scaling of Transformer self-attention has driven the adoption of sub-quadratic Selective State Space Models (SSMs) like Mamba, which compress past context into a fixed-size recurrent hidden state. This strict informational bottleneck raises a foundational question for mechanistic interpretability: do SSMs and Transformers learn fundamentally distinct latent representations? In this work, we employ Sparse Autoencoders (SAEs) to...

    arxiv.org/abs/2609.24440 · PDF

  32. 32

    1% of Tokens Can Be Enough: On Gradient Estimation in On-Policy Distillation

    Huanxin Sheng, Zhiling Ye, Haonan Wang, Jian Wang, Jinjie Gu, Jian Kang

    cs.LG · cs.CL

    Sparse on-policy distillation (OPD) allocates teacher supervision to a small subset of tokens in student-generated trajectories. However, useful teacher guidance can yield a noisy update when its gradient is estimated from a sampled next token. We study this estimation problem at a fixed prefix in information geometry and propose an information-efficiency ratio (IER) based on a signal-to-noise decomposition. IER characterizes relative...

    arxiv.org/abs/2609.24432 · PDF

  33. 33

    Prior-Amortized In-Context Bayesian Inference for Generalized Linear Mixed-Effects Models

    Alex Kipnis, Marcel Binz, Eric Schulz

    cs.LG · stat.ML

    Hierarchical data is ubiquitous in the empirical sciences and is most commonly analyzed with generalized linear mixed-effects models (GLMMs). Bayesian inference for GLMMs yields calibrated uncertainty but requires MCMC; the No-U-Turn Sampler (NUTS) is the gold standard but is slow and must restart from scratch for every new dataset, model and prior. We introduce metabeta, a pretrained neural network for prior-amortized in-context Bayesian...

    arxiv.org/abs/2609.24422 · PDF

  34. 34

    ARM: Attention with Routed-Memory for Learnable Sparse Control

    Qiuhao Zeng, Jerry Huang, Peng Lu, Ruiyi Fang, Gezheng Xu, Zihao Jing, Yufei Cui, Charles Ling, Gang Niu, Boyu Wang

    cs.LG · cs.AI

    Despite advances in long-context inference, large language models (LLMs) remain fundamentally limited by the key-value (KV) caching mechanisms that are necessary for stable computation. Techniques such as selective token eviction and pruning have vastly mitigated these issues, but often discard core information to manage the growing cache. In this paper, we propose Attention with Routed Memory (ARM) a novel KV caching structure that...

    arxiv.org/abs/2609.24417 · PDF

  35. 35

    Artificial Structure Function Search: Preserving Artificial Functional Connectivity for Structured Pruning

    Mindula Illeperuma, Rafael Pina, Charuka Herath, Sharmarke A. Gabayre, Varuna De Silva

    cs.LG · cs.AI

    Structured pruning is a model compression technique that is used to reduce the computational cost of deploying deep neural networks on resource-constrained devices. Popular methods of pruning rely on opaque heuristics or weight-based criteria that give no indication as to the structural dependencies in the network. To address these limitations we present Artificial Structure Function Search (ASF-S): a novel structured pruning framework. ASF-S...

    arxiv.org/abs/2609.24401 · PDF

  36. 36

    Probabilistic Modelling of Operational Design Domains, A New Approach for Testing AI Systems

    Hans-Werner Wiesbrock

    cs.LG

    The conventional testing process quickly fails when applied to ML-based systems such as obstacle detection in vehicles: if an obstacle is not detected in a test, classical bug fixing is impossible and an AI system will always retain shortcomings. Test results can therefore only be interpreted statistically, which in turn requires test sets that are not only complete with respect to the operational design domain (ODD) of the system, but also...

    arxiv.org/abs/2609.24397 · PDF

  37. 37

    Climate Variability Modulates the Impact of Price Spikes on Food Insecurity

    Jordi Cerdà-Bautista, Vasileios Sitokonstantinou, Homer Durand, Gherardo Varando, Michele Ronco, Gustau Camps-Valls

    cs.LG

    Climate variability influences whether a market disruption escalates into a food crisis, yet broad climate patterns like El Niño, tracked months before they alter hydro-climatic conditions, are still not incorporated as an early-warning component in food-security responses. We address this gap by introducing sensitivity regimes, a stratification of regions by the direction and strength of their vegetation response to the El Niño Southern...

    arxiv.org/abs/2609.24394 · PDF

  38. 38

    NAVIR: Neuromorphic Audio-Visual Speech Recognition for Robust Human-Robot Interaction on Edge Hardware

    Leonidas Delimpasis, Panagiota Moraiti, Antonis Porichis, Panos Chatzakos, Michail Karamousadakis

    cs.LG

    Voice-controlled interaction in industrial settings is hampered by acoustic noise, which severely degrades audio-only speech recognition. Audio-visual speech recognition (AVSR) addresses this by fusing lip-motion cues with the audio stream, but state-of-the-art pipelines rely on three-dimensional convolutions, recurrent units, and attention modules that exceed the budget of typical edge devices. We present NAVIR, an end-to-end AVSR system...

    arxiv.org/abs/2609.24391 · PDF

  39. 39

    Machine Learning-Based Prediction of Childhood Stunting in Bangladesh: Fairness and Temporal Robustness Assessment

    Md Ahshanul Haque, Muhammad Ashad Kabir

    cs.LG

    Childhood stunting remains a major public health concern in Bangladesh and reflects long-term growth failure influenced by child, maternal, household, socioeconomic, and health-service factors. This study used nationally representative Bangladesh Demographic and Health Survey data from 2007 to 2022 to develop machine learning models for population-level prediction of childhood stunting and to assess temporal robustness and subgroup fairness....

    arxiv.org/abs/2609.24386 · PDF

  40. 40

    Credit Access is Associated with Improved Food Security in the Horn of Africa

    Jordi Cerdà-Bautista, Vasileios Sitokonstantinou, José Manuel Veiga López-Peña, Duccio Piovani, José María Tárraga,...

    cs.LG

    The intensification of climate change poses a growing threat to food security, especially in vulnerable communities. This study employs an observational machine-learning framework to estimate the causal association between access to credit and acute food insecurity in Somalia and across the Horn of Africa, drawing on a harmonized dataset spanning key environmental, socioeconomic, and conflict-related factors from 2015 to 2022. Results...

    arxiv.org/abs/2609.24382 · PDF

  41. 41

    Information-Time Proximal Policy Optimization

    Yongcheng Zeng, Xinyu Cui, Yan Song, Guoqing Liu, Hongsheng Xin, Kaike Zhang, Cheng Deng, Kun Zhan, Jian Ying, Jian...

    cs.LG · cs.AI

    RLVR has substantially improved the reasoning capabilities of LLMs. However, existing methods typically parameterize temporal progression in the Markov Decision Process by token-by-token generation, despite the highly non-uniform information flow along autoregressive trajectories. In this paper, we propose InfoPPO, which reparameterizes temporal progression using information density rather than raw token count. This reparameterization induces...

    arxiv.org/abs/2609.24380 · PDF

  42. 42

    Prescriptive SVD-Inspired Attention via Spectral Energy Retention

    Vasileios Arampatzakis, Vasileios Sevetlidis, George Pavlidis

    cs.LG · cs.CV

    Self-attention is central to modern Transformer architectures, but its dense dot-product formulation makes it difficult to identify which internal directions are structurally important and which can be modified without disrupting the model. SVD-Inspired Attention (SVDA) addresses part of this problem by introducing a learned diagonal spectrum into the query-key score interaction, making latent attention directions explicitly inspectable...

    arxiv.org/abs/2609.24370 · PDF

  43. 43

    Explainable Neuro-Fuzzy Prediction for Trustworthy Decision-Making in Maritime

    Dionisis Kalogeropoulos, Georgia Sovatzidi, Dimitris K. Iakovidis

    cs.LG

    Predicting when maritime systems require maintenance can be critical, avoiding hazards and costly consequences. To address this problem, this paper proposes an explainable decision-making framework that integrates a neuro-fuzzy prediction model with a two-stage explainable component. The first stage of this component produces feature-attribution explanations, using gradient-based saliency maps, and the second stage extracts local rules using...

    arxiv.org/abs/2609.24358 · PDF

  44. 44

    Pharmacokinetic State Space Models for Unbiased Prediction of Haemodynamic Collapse

    Rithin Nagaraj, Sudiksha Chindula, Bhaskarjyoti Das

    cs.LG

    An Intraoperative Hypotension (IOH) event is a frequent complication during administration of general anaesthesia with serious downstream consequences, yet clinical management remains reactive and not predictive. Existing predictive models, however, ignore drug infusion history as a valuable signal for prediction despite its direct pharmacological relevance. Our model achieves an Area Under the Receiver Operating Characteristic curve (AUROC)...

    arxiv.org/abs/2609.24338 · PDF

  45. 45

    A Distributional Optimisation Perspective on Combining Models in Deep Learning

    Congye Wang, Yan Lin, Zheyang Shen, Matthew A. Fisher, Chris. J. Oates

    cs.LG · stat.ML

    Combining predictions from different models can improve performance at machine learning tasks, but the training of the individual models and the rule used to combine them are typically chosen separately, and by ad hoc means. Recent advances in distributional optimisation (i.e. where the optimisation occurs over the set of probability distributions) offer an opportunity for principled joint training, viewing the collection of models as a...

    arxiv.org/abs/2609.24328 · PDF

  46. 46

    The Undetected Damage of Quantization on Retrieval and How to Fix It

    Luca Zhou, Alessandro Zirilli, Daniele Solombrino, Roberto Dessì, Emanuele Rodolà

    cs.LG · cs.AI

    We show that a quantized model that keeps its classification accuracy still changes $14$ to $46\%$ of its top-1 retrieval results, and that aggregate ranking metrics reveal only part of this damage. We tie this failure to the gap between the two highest scores and use that gap to decide when a quantized answer can be trusted and where additional precision should be spent. We show that the top-1 result is guaranteed to survive quantization...

    arxiv.org/abs/2609.24322 · PDF

  47. 47

    SupportCal: Label-Free Calibration of Post-Trained LLMs via Reference Support and Corroboration

    Linhan Luo, Lequan Lin, Dai Shi, Feng Chen, José Miguel Hernández-Lobato, Junbin Gao

    cs.LG · cs.CL

    Post-training often improves task performance but can degrade confidence calibration, leaving post-trained language models (PoLMs) more overconfident than their corresponding pretrained language models (PLMs). Because task-specific labeled calibration data can be costly or unavailable, the corresponding pretrained PLM provides a natural label-free reference for post-hoc calibration. Prior agreement-gated PLM-referenced calibration fits a...

    arxiv.org/abs/2609.24303 · PDF

  48. 48

    KV-COBRA: KV Cache Compression via Co-Optimized Bit-Rank Allocation

    Sihyeon Ha, Jaeho Lee, Yo-Seb Jeon

    cs.LG · cs.AI

    What limits KV-cache compression at extreme bit-rates? We argue that it is not the choice of compression scheme, but how its budget is allocated across attention heads. Existing methods apply rank and bit-width uniformly, ignoring that each head has a different optimal mix of rank truncation and quantization. We show that co-optimizing rank and bit-width per head, using only standard low-rank projection and scalar quantization, dominates...

    arxiv.org/abs/2609.24298 · PDF

  49. 49

    TTSE: A Two-Track Online Self-Evolution Framework

    Ruimin Pei, Yongkang Wu, Shangyi Zheng, Yaqing Zhang, Deyang Li, Jianjun Tao, Xinyu Zhang, Xiang Zhang

    cs.LG · cs.AI

    As Large Language Model (LLM) agents are applied in continuously interactive environments, driving the evolution of their own capabilities becomes a core problem for achieving long-term autonomy. Currently, environmental knowledge is typically treated as an external fixed input rather than as part of the agent's ongoing evolution. Reinforcement learning methods usually optimize policies through environmental interaction but tend to adapt only...

    arxiv.org/abs/2609.24289 · PDF

  50. 50

    High-Dimensional Online Change Point Detection with Adaptive Thresholding and Interpretability

    Sven Jacob, Bardh Prenkaj, Weijia Shao, Gjergji Kasneci

    cs.LG

    Change point detection (CPD) identifies abrupt and significant changes in sequential data, with applications in human activity recognition, financial markets, cybersecurity, manufacturing, and autonomous systems. Traditional CPD methods often face computational challenges in high-dimensional settings and typically provide limited explanations for detected changes, which can restrict their practical usability. This paper introduces a CPD...

    arxiv.org/abs/2609.24278 · PDF

  51. 51

    MemCalib: Benchmarking and Optimizing Memory Use in LLM Agents

    Ruike Cao, Fanyu Zhao, Fugen Yao, Liang Dong, Jian Xu, Guanjun Jiang, Yifei Zhao, Han Zhang, Li Xiao

    cs.LG · cs.AI

    The effectiveness of agent memory ultimately depends on whether the underlying LLM gives each memory in context an appropriate degree of influence over its response. Yet this capability has remained largely overlooked. To assess this capability, we introduce MemCalib, a benchmark grounded in realistic memory-system scenarios for evaluating memory use and advancing optimization algorithms. Results on the MemCalib test set reveal that frontier...

    arxiv.org/abs/2609.24259 · PDF

  52. 52

    Explainable Predictive Condition-based Maintenance of Naval-Propulsion Systems using Fuzzy Logic

    Dionisis Kalogeropoulos, Georgia Sovatzidi, Panagiotis G. Kalozoumis, Dimitris K. Iakovidis

    cs.LG

    The shipping industry has a significant impact on the global economy, emphasizing the need for operational availability and safety through the use of effective maintenance techniques. During the last decades, predictive maintenance (PdM) has emerged as a promising solution compared to the existing conventional maintenance systems. This is because it offers several advantageous functions, such as damage predictions for vessel components,...

    arxiv.org/abs/2609.24250 · PDF

  53. 53

    Reinforcement Learning Inspired Black-box Adversarial Attacks for Computer Vision

    Florian Krone, Elena Hoemann, Sven Hallerbach

    cs.LG · cs.CR · cs.CV

    Neural networks, both convolution or transformer based, are essential for modern computer vision systems. However, they are vulnerable to small perturbations, almost imperceptible to humans, which significantly alter the model's prediction. These adversarial attacks are often considered to be a significant threat to the implementation of neural networks in safety-critical applications. Most attacks utilize the white-box threat model and...

    arxiv.org/abs/2609.24249 · PDF

  54. 54

    Hessian Rank Constraint for Learning Structure of Nonlinear Latent Variable Models

    Zijian Li, Ruichu Cai, Feng Xie, Xinshuai Dong, Haoyue Dai, Yuewen Sun, Yujia Zheng, Guangyi Chen, Yingyao Hu, Kun Zhang

    cs.LG · cs.AI

    Uncovering latent variables and their causal relations from observed data is a fundamental yet challenging problem. Existing methods often rely on restrictive assumptions, such as linear relations or invertible mixing functions. To better address this problem under general nonlinear mixing procedures, we propose a condition called the cross-Hessian Rank Constraint (HRC), which serves as a primitive rank-based tool for nonlinear latent causal...

    arxiv.org/abs/2609.24241 · PDF

  55. 55

    Adaptive Forgetting for Nonstationary Optimization: Towards Robust EEG Decoding

    Hongyu Zhu, Lin Chen, Jing Chen, Yuting Zhou, Mingsheng Shang

    cs.LG · cs.HC

    Electroencephalography (EEG) provides non-invasive monitoring of brain activity and is widely used in emotion recognition, motor imagery and sleep staging. Although within-subject decoding has achieved considerable progress, cross-subject generalization remains a central challenge in practical applications. EEG decoders are typically trained with Adam/AdamW under a fixed second-moment decay coefficient, even though cross-subject learning...

    arxiv.org/abs/2609.24233 · PDF

  56. 56

    Displacement Geometry Captures Platonic Shared Reality Across Models and Modalities

    Chenming Shang, Yujin Tang, Jun Jie Ou Yang, Ruize Xu, Adam Breuer, Nikhil Singh

    cs.LG

    The Platonic Representation Hypothesis (PRH) claims that independently trained models converge on a shared statistical model of reality, yet recent work finds only weak pointwise similarity between models. In this paper, we show that what models share is not the location of samples in representation space, but the directions (displacement vectors) between them. Under a single orthogonal alignment--rotation and reflection only--these...

    arxiv.org/abs/2609.24209 · PDF

  57. 57

    Opinion Leader Dynamics: How Sparse Attention Shapes Token Clustering

    Jingkun Liu, Yue Song

    cs.LG · cs.AI

    Sparse attention reduces the quadratic cost of global self-attention while retaining strong empirical performance, but how its restricted interactions shape the evolution of token representations remains theoretically underexplored. Modeling tokens as particles on the unit sphere, we introduce opinion leader dynamics, a framework that identifies two mechanisms through which token groups converge internally while maintaining distinct limiting...

    arxiv.org/abs/2609.24202 · PDF

  58. 58

    H-Spec: Parallel Speculative Decoding Without a Drafter-Side KV Cache

    Weifan Jiang, Krishna Teja Chitty-Venkata, Megan Flynn, Reed Meyerson, Zhenting Qi, Tianyu Wu, Eldar Kurtic, Minlan...

    cs.LG

    Speculative decoding losslessly accelerates large language model inference by having a lightweight draft model predict future tokens for verification by the target model. Recent block diffusion drafters further reduce drafting latency by predicting multiple tokens in parallel. However, existing block drafters project target hidden states at every input position into a separate drafter-side KV cache, incurring per-request memory and KV-write...

    arxiv.org/abs/2609.24197 · PDF

  59. 59

    Acceptance-Aware Draft Model Training for Speculative Decoding

    Tianhua Xia, Mugilan Ganesan, Yifei Feng, Haiyu Wang, Maximilian Egger, Sai Qian Zhang

    cs.LG

    Speculative decoding accelerates large language model (LLM) inference by using a lightweight draft model to generate multiple candidate tokens that are verified by the target model in a single forward pass. Its speedup is largely determined by the acceptance length, yet existing draft-model training methods mainly optimize cross-entropy or Kullback-Leibler (KL) divergence as proxies. These objectives encourage distribution matching but do not...

    arxiv.org/abs/2609.24150 · PDF

  60. 60

    Mind or Message? Auditing Theory of Mind in Multi-Agent Social Simulation

    Cong Li, Cheng Chen, Thomas Fung, Alex Rossi, Yi Li

    cs.LG

    Language model agents are increasingly used to simulate social interaction, and the resulting transcripts read as though the agents understand one another. We ask whether that appearance rests on a model of the partner's mind or on the surface record of what the partner said. We build a social simulation in which both questions have exact answers: 40 multi-issue negotiations whose hidden preference weights and whose full Pareto frontier are...

    arxiv.org/abs/2609.24146 · PDF

  61. 61

    Luck Is Not Skill: When Do Paired Rollouts Help Group-Relative RL of LLM Agents?

    Nazmus Sakib

    cs.LG

    Group-relative reinforcement learning compares rollouts of the same prompt, but independent environment noise can obscure these comparisons. We study paired rollouts, which share an event-keyed noise schedule within each group while preserving each rollout's marginal distribution. Pairing removes the between-schedule component of reward-contrast variance, but need not reduce gradient variance. For one-sided grader noise, we derive an exact...

    arxiv.org/abs/2609.24144 · PDF

  62. 62

    CLOOPD: Closing the Learner Loop in On-Policy Distillation

    Keye Zheng, Hanyu Li, Zhan Cheng, Yuan Gao

    cs.LG

    On-policy distillation (OPD) pays twice for each fresh batch: the student generates trajectories and a stronger teacher scores them. Existing methods improve which trajectories are scored and how the teacher signal is constructed, but usually consume it with one actor update. We introduce CLOOPD, a closed-loop framework separating teacher-signal acquisition from student-side realization. CLOOPD selects an adaptive $α$ waypoint inside a KL...

    arxiv.org/abs/2609.24141 · PDF

  63. 63

    PAC-Bayesian Meta-Learning for Few-Shot Identification of Linear Dynamical Systems

    Chenfeng Huang, George Michailidis

    cs.LG

    Identifying linear time-invariant (LTI) dynamical systems is challenging when trajectories are short, noisy, or high-dimensional. Traditional system identification typically treats each system independently and cannot exploit shared structure across related systems. We propose PBML-LTI, a PAC-Bayesian meta-learning framework for few-shot LTI system identification that learns a transferable prior over task-specific dynamics while preserving...

    arxiv.org/abs/2609.24117 · PDF

  64. 64

    SPeaR: Test-Time Adaptation with Steering Primitives for Realigning Representations

    Muhammad Sudipto Siam Dip, Ali Etemad

    cs.LG · cs.CV

    Test-time adaptation (TTA) addresses distribution shift using only unlabeled test data. Existing methods typically adapt pretrained models by updating their parameters, limiting both what is adapted and where adaptation can occur within the network. We instead keep the pretrained network frozen and steer its intermediate representations. We introduce SPeaR (Steering Primitive for Realigning Representations), which inserts lightweight...

    arxiv.org/abs/2609.24111 · PDF

  65. 65

    Reinforcement Learning under State and Outcome Uncertainty: A Foundational Distributional Perspective

    Larry Preuett, Qiuyi Zhang, Muhammad Aurangzeb Ahmad

    cs.LG · cs.AI

    In many real-world planning tasks, agents must tackle uncertainty about the environment's state and variability in the outcomes of any chosen policy. We address both forms of uncertainty as a first step toward safer algorithms in partially observable settings. Specifically, we extend Distributional Reinforcement Learning (DistRL)-which models the entire return distribution for fully observable domains-to Partially Observable Markov Decision...

    arxiv.org/abs/2609.24103 · PDF

  66. 66

    FlashBoB: I/O-Efficient Exact Backward-over-Backward for Softmax Attention

    Anthony Givans, Michael Crawshaw, Mingrui Liu

    cs.LG · cs.AI

    Transformer models built on the attention mechanism have become a central building block in modern deep learning, yet softmax attention remains a major bottleneck for long-context workloads. While FlashAttention makes the forward and first backward passes I/O-efficient, it does not support backward-over-backward (BoB), which enables exact differentiation through the backward pass for applications such as second-order optimization, test-time...

    arxiv.org/abs/2609.24089 · 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.