cs.LG papers | Gist.Science

Autocorrelation effects in a stochastic-process model for decision making via time series

This study employs a stochastic-process model to demonstrate that the optimal autocorrelation of time-series signals for solving multi-armed bandit problems depends on the reward environment, with negative autocorrelation being advantageous in reward-rich settings and positive autocorrelation in reward-poor ones, while performance remains independent of autocorrelation when the sum of winning probabilities equals one.

Tomoki Yamagami, Mikio Hasegawa, Takatomo Mihana, Ryoichi Horisaki, Atsushi Uchida2026-03-09🔬 physics.optics

Towards Efficient and Stable Ocean State Forecasting: A Continuous-Time Koopman Approach

This paper demonstrates that the Continuous-Time Koopman Autoencoder (CT-KAE) serves as a lightweight, stable, and efficient surrogate model for long-horizon ocean state forecasting, outperforming autoregressive Transformer baselines by maintaining bounded errors and consistent large-scale statistics over 2083-day rollouts while enabling resolution-invariant predictions.

Rares Grozavescu, Pengyu Zhang, Mark Girolami, Etienne Meunier2026-03-09🔬 physics.app-ph

When AI Levels the Playing Field: Skill Homogenization, Asset Concentration, and Two Regimes of Inequality

This paper presents a task-based model demonstrating that while generative AI homogenizes individual skills, it can simultaneously increase aggregate inequality by shifting economic value toward concentrated complementary assets, with the net outcome determined by the technology's structure and labor market institutions rather than a single universal verdict.

Xupeng Chen, Shuchen Meng2026-03-09🤖 cs.AI

Aligning the True Semantics: Constrained Decoupling and Distribution Sampling for Cross-Modal Alignment

This paper proposes CDDS, a novel cross-modal alignment algorithm that utilizes a dual-path UNet for constrained decoupling of semantic and modality components and a distribution sampling method to bridge the modality gap, thereby achieving superior semantic consistency and outperforming state-of-the-art methods by 6.6% to 14.2%.

Xiang Ma, Lexin Fang, Litian Xu, Caiming Zhang2026-03-09🤖 cs.LG

FuseDiff: Symmetry-Preserving Joint Diffusion for Dual-Target Structure-Based Drug Design

FuseDiff is a novel end-to-end diffusion model that jointly generates a single ligand and two pocket-specific binding poses for dual-target structure-based drug design by employing a symmetry-preserving message-passing backbone with Dual-target Local Context Fusion to overcome the limitations of existing staged pipelines.

Jianliang Wu, Anjie Qiao, Zhen Wang, Zhewei Wei, Sheng Chen2026-03-09🤖 cs.LG

Learning Optimal Distributionally Robust Individualized Treatment Rules Integrating Multi-Source Data

This paper proposes a prior information-based distributionally robust individualized treatment rule (PDRO-ITR) that integrates multi-source data to address posterior shift by maximizing the worst-case policy value over a covariate-dependent uncertainty set, thereby ensuring robust performance and achieving superior results in simulations and real-world applications.

Wenhai Cui, Wen Su, Xingqiu Zhao2026-03-09🤖 cs.LG

Machine Learning for analysis of Multiple Sclerosis cross-tissue bulk and single-cell transcriptomics data

This study presents an end-to-end machine learning pipeline utilizing XGBoost and SHAP explainability to integrate bulk and single-cell transcriptomic data from multiple sclerosis patients, successfully identifying high-performance biomarkers and novel mechanistic pathways involving immune activation, non-canonical checkpoints, and Epstein-Barr virus-related processes.

Francesco Massafra, Samuele Punzo, Silvia Giulia Galfré, Alessandro Maglione, Simone Pernice, Stefano Forti, Simona Rolla, Marco Beccuti, Marinella Clerico, Corrado Priami, Alina Sîrbu2026-03-09🤖 cs.LG

Why Depth Matters in Parallelizable Sequence Models: A Lie Algebraic View

This paper utilizes a Lie-algebraic control perspective to demonstrate that increasing the depth of parallelizable sequence models exponentially reduces approximation error by expanding their expressivity through a tower of Lie algebra extensions, a finding validated by experiments on symbolic and continuous state-tracking tasks.

Gyuryang Heo, Timothy Ngotiaoco, Kazuki Irie, Samuel J. Gershman, Bernardo Sabatini2026-03-09🤖 cs.LG

Prediction-Powered Conditional Inference

This paper proposes a prediction-powered conditional inference method that leverages abundant unlabeled data and machine learning predictors to construct valid, variance-reduced confidence intervals for conditional functionals at fixed test points without requiring parametric models, while establishing rigorous nonasymptotic error bounds and asymptotic normality.

Yang Sui, Jin Zhou, Hua Zhou, Xiaowu Dai2026-03-09🤖 cs.LG

Koopman Regularized Deep Speech Disentanglement for Speaker Verification

This paper introduces the Deep Koopman Speech Disentanglement Autoencoder (DKSD-AE), a scalable and efficient architecture that leverages Koopman operators and instance normalization to effectively disentangle speaker identity from linguistic content for robust speaker verification without relying on textual supervision or large pretrained models.

Nikos Chazaridis, Mohammad Belal, Rafael Mestre, Timothy J. Norman, Christine Evers2026-03-09🤖 cs.LG

A Novel Hybrid Heuristic-Reinforcement Learning Optimization Approach for a Class of Railcar Shunting Problems

This paper proposes a novel Hybrid Heuristic-Reinforcement Learning (HHRL) framework that integrates railway-specific heuristics with Q-learning to efficiently solve complex railcar shunting problems involving both one-sided and two-sided classification tracks by decomposing multi-locomotive tasks into manageable subproblems.

Ruonan Zhao, Joseph Geunes2026-03-09🤖 cs.LG

Spatiotemporal Heterogeneity of AI-Driven Traffic Flow Patterns and Land Use Interaction: A GeoAI-Based Analysis of Multimodal Urban Mobility

This study proposes and validates a GeoAI hybrid framework integrating MGWR, Random Forest, and ST-GCN to effectively model the spatiotemporal heterogeneity of multimodal traffic flows and their interaction with land use, demonstrating superior predictive accuracy and revealing distinct urban traffic typologies that underscore the critical role of local morphological context in mobility planning.

Olaf Yunus Laitinen Imanov2026-03-09🤖 cs.AI

Bias In, Bias Out? Finding Unbiased Subnetworks in Vanilla Models

This paper introduces Bias-Invariant Subnetwork Extraction (BISE), a method that identifies and isolates fair, bias-agnostic subnetworks within standard pre-trained models through pruning, enabling effective bias mitigation without retraining or additional unbiased data.

Ivan Luiz De Moura Matos, Abdel Djalil Sad Saoud, Ekaterina Iakovleva, Vito Paolo Pastore, Enzo Tartaglione2026-03-09🤖 cs.LG

On the Value of Tokeniser Pretraining in Physics Foundation Models

This paper demonstrates that pretraining tokenizers with an autoencoding objective before training dynamics models significantly enhances the computational efficiency and accuracy of physics foundation models, particularly when the pretraining data aligns with the downstream physical system.

Hadi Sotoudeh, Payel Mukhopadhyay, Ruben Ohana, Michael McCabe, Neil D. Lawrence, Shirley Ho, Miles Cranmer2026-03-09🔭 astro-ph

From Decoupled to Coupled: Robustness Verification for Learning-based Keypoint Detection with Joint Specifications

This paper introduces the first coupled robustness verification framework for heatmap-based keypoint detectors that uses a mixed-integer linear program to jointly bound deviations across all keypoints, thereby providing sound and less conservative guarantees than prior decoupled methods.

Xusheng Luo, Changliu Liu2026-03-09🤖 cs.LG

Behavior-dLDS: A decomposed linear dynamical systems model for neural activity partially constrained by behavior

This paper introduces behavior-decomposed linear dynamical systems (b-dLDS), a novel modeling approach that disentangles behavior-related neural dynamics from internal computations in large-scale brain recordings, demonstrating superior performance over existing supervised models and successfully scaling to tens of thousands of neurons in zebrafish hindbrain data.

Eva Yezerets, En Yang, Misha B. Ahrens, Adam S. Charles2026-03-09🤖 cs.LG

RACAS: Controlling Diverse Robots With a Single Agentic System

The paper introduces RACAS, a robot-agnostic agentic system that uses natural language communication between LLM/VLM-based modules to control diverse robotic platforms without requiring code modifications or retraining, successfully demonstrating its effectiveness across wheeled, multi-jointed, and underwater robots.

Dylan R. Ashley, Jan Przepióra, Yimeng Chen, Ali Abualsaud, Nurzhan Yesmagambet, Shinkyu Park, Eric Feron, Jürgen Schmidhuber2026-03-09🤖 cs.AI

Identifying Adversary Characteristics from an Observed Attack

This paper proposes a domain-agnostic framework to identify the most probable characteristics of an attacker from an observed data-manipulation attack, demonstrating that such identification enables more effective exogenous mitigation and improves the performance of learning-based defenses.

Soyon Choi, Scott Alfeld, Meiyi Ma2026-03-09🤖 cs.LG

Making Reconstruction FID Predictive of Diffusion Generation FID

This paper introduces interpolated FID (iFID), a novel metric that achieves a strong correlation with diffusion generation FID by interpolating latent representations between dataset samples and their nearest neighbors, thereby overcoming the limitations of traditional reconstruction FID in predicting generative model quality.

Tongda Xu, Mingwei He, Shady Abu-Hussein, Jose Miguel Hernandez-Lobato, Haotian Zhang, Kai Zhao, Chao Zhou, Ya-Qin Zhang, Yan Wang2026-03-09🤖 cs.LG

When Rubrics Fail: Error Enumeration as Reward in Reference-Free RL Post-Training for Virtual Try-On

This paper introduces Implicit Error Counting (IEC), a reference-free reinforcement learning post-training method that enumerates and weights errors to generate rewards, demonstrating superior performance over Rubrics as Rewards (RaR) in virtual try-on tasks where multiple valid outputs exist and ideal reference answers are unavailable.

Wisdom Ikezogwo, Mehmet Saygin Seyfioglu, Ranjay Krishna, Karim Bouyarmane2026-03-09🤖 cs.AI

← Previous Next →