Ax Zihan Wang, Chi Gui, Xing Jin, Qineng Wang, Licheng Liu, Kangrui Wang, Shiqi Chen, Linjie Li, Zhengyuan Yang, Pingyue Zhang, Yiping Lu, Jiajun Wu, Li Fei-Fei, Lijuan Wang, Yejin Choi, Manling Li 21d ago

RAGEN-2: Reasoning Collapse in Agentic RL

RAGEN-2 identifies reasoning collapse in RL-trained multi-turn LLM agents where models use input-agnostic templates despite stable entropy metrics.

Ax Giulia Bertaglia, Raffaella Fiamma Cabini 21d ago

Asymptotic-Preserving Neural Networks for Viscoelastic Parameter Identification in Multiscale Blood Flow Modeling

Neural networks for identifying viscoelastic parameters in multiscale blood flow cardiovascular models using asymptotic-preserving methods.

Ax Lin Mu, Haiyang Wang, Li Ni, Lei Sang, Zhize Wu, Peiquan Jin, Yiwen Zhang 21d ago

TalkLoRA: Communication-Aware Mixture of Low-Rank Adaptation for Large Language Models

TalkLoRA: communication-aware mixture of LoRA experts for parameter-efficient LLM fine-tuning addressing routing instability in MoE-augmented approaches.

Ax Wenyue Hua, Sripad Karne, Qian Xie, Armaan Agrawal, Nikos Pagonas, Kostis Kaffes, Tianyi Peng 21d ago

AgentOpt v0.1 Technical Report: Client-Side Optimization for LLM-Based Agent

AgentOpt framework for client-side optimization of LLM-based agents handling composition of local tools, remote APIs and diverse models with reduced costs.

Ax Suraj Yadav, Siddharth Yadav, Parth Goyal 21d ago

Limits of Difficulty Scaling: Hard Samples Yield Diminishing Returns in GRPO-Tuned SLMs

GRPO preference optimization applied to small language models shows diminishing returns on hard samples, revealing capacity boundaries in math reasoning tasks.

Ax Yi Yang, Ovidiu Daescu 21d ago

BiScale-GTR: Fragment-Aware Graph Transformers for Multi-Scale Molecular Representation Learning

Graph Transformer architecture combining GNNs and Transformers for multi-scale molecular property prediction with fragment-aware representation learning.

Ax Marzi Heidari, Hanping Zhang, Hao Yan, Yuhong Guo 21d ago

Bi-Level Optimization for Single Domain Generalization

Bi-level optimization framework (BiSDG) for single domain generalization that decouples task learning from domain modeling using surrogate distributions.

Ax Rishab Balasubramanian, Pin-Jie Lin, Rituraj Sharma, Anjie Fang, Fardin Abdi, Viktor Rozgic, Zheng Du, Mohit Bansal, Tu Vu 21d ago

The Master Key Hypothesis: Unlocking Cross-Model Capability Transfer via Linear Subspace Alignment

Master Key Hypothesis proposes capabilities correspond to transferable directions in low-dimensional subspace; introduces UNLOCK for training-free cross-model capability transfer.

Ax Sakib Mostafa, Lei Xing, Md. Tauhidul Islam 21d ago

Toward a universal foundation model for graph-structured data

Develops universal foundation model for graph-structured biomedical data including molecular networks and regulatory circuits.

Ax Ruggero Freddi, Nicolas Seseri, Diana Nigrisoli, Alessio Basti 21d ago

Bridging Theory and Practice in Crafting Robust Spiking Reservoirs

Addresses hyperparameter tuning challenges in spiking reservoir computing by introducing robustness interval concept for edge-of-chaos operation.

Ax Xiao Shou 21d ago

ODE-free Neural Flow Matching for One-Step Generative Modeling

OT-NFM enables one-step generative modeling by learning transport maps directly instead of integrating vector fields, achieving single forward pass generation.

Ax Mingchen Zhuge, Changsheng Zhao, Haozhe Liu, Zijian Zhou, Shuming Liu, Wenyi Wang, Ernie Chang, Gael Le Lan, Junjie Fei, Wenxuan Zhang, Yasheng Sun, Zhipeng Cai, Zechun Liu, Yunyang Xiong, Yining Yang, Yuandong Tian, Yangyang Shi, Vikas Chandra, J\"urgen Schmidhuber 21d ago

Neural Computers

Proposes Neural Computers (NCs) that unify computation, memory, and I/O in learned runtime states, aiming toward fully neural computing systems that replace explicit programs.

Ax Yi Xu, Philipp Jettkant, Laura Ruis 21d ago

The Depth Ceiling: On the Limits of Large Language Models in Discovering Latent Planning

Study on latent reasoning limits in LLMs investigating whether models discover multi-step planning strategies without supervision.

Ax Srinidhi Madabhushi, Pranesh Vyas, Swathi Vaidyanathan, Mayur Kurup, Elliott Nash, Yegor Silyutin 21d ago

From Load Tests to Live Streams: Graph Embedding-Based Anomaly Detection in Microservice Architectures

Graph embedding-based anomaly detection for microservice architectures identifying under-represented services in load testing.

Ax Noufa Haneefa, Teddy Lazebnik, Einav Peretz-Andersson 21d ago

Quality-preserving Model for Electronics Production Quality Tests Reduction

Data-driven approach reducing electronics production test costs while adapting to changing defect distributions and controlling escape risk.

Ax Yuanjie Shi, Peihong Li, Zijian Zhang, Janardhan Rao Doppa, Yan Yan 21d ago

Conformal Margin Risk Minimization: An Envelope Framework for Robust Learning under Label Noise

Conformal Margin Risk Minimization framework for robust classification under label noise without privileged knowledge.

Ax Willa Potosnak, Nina \.Zukowska, Micha{\l} Wili\'nski, Dan Howarth, Ignacy St\k{e}pka, Mononito Goswami, Artur Dubrawski 21d ago

MICA: Multivariate Infini Compressive Attention for Time Series Forecasting

MICA architecture for multivariate time series forecasting addressing Transformer scalability with channel-dependent attention.

Ax Dev Arpan Desai, Shaoyi Huang, Zining Zhu 21d ago

Distributed Interpretability and Control for Large Language Models

Multi-GPU implementation of activation-level interpretability and steering techniques for large language models, extending single-GPU methods to distributed settings.

Ax David Cho, Yifan Wang, Fanping Sui, Ananth Grama 21d ago

Inference-Time Code Selection via Symbolic Equivalence Partitioning

Novel inference-time scaling method using symbolic execution to select correct code generation solutions from LLM candidates without expensive external verifiers.

Ax Maojiang Su, Po-Chung Hsieh, Weimin Wu, Mingcheng Lu, Jiunhau Chen, Jerry Yao-Chieh Hu, Han Liu 21d ago

Discrete Flow Matching Policy Optimization

DoMinO: unified RL framework for fine-tuning discrete flow matching models viewing sampling as multi-step MDP.

Ax Andrew Lowy 21d ago

Optimal Rates for Pure {\varepsilon}-Differentially Private Stochastic Convex Optimization with Heavy Tails

Theoretical analysis of stochastic convex optimization with heavy-tailed gradients under differential privacy constraints.

Ax Philipp Hellwig, Willem Zuidema, Claire E. Stevenson, Martha Lewis 21d ago

Transformer See, Transformer Do: Copying as an Intermediate Step in Learning Analogical Reasoning

Study of transformers learning analogical reasoning via copying intermediate representations using meta-learning for compositionality.

Ax Peigui Qi, Kunsheng Tang, Yanpu Yu, Jialin Wu, Yide Song, Wenbo Zhou, Zhicong Huang, Cheng Hong, Weiming Zhang, Nenghai Yu 21d ago

VLMShield: Efficient and Robust Defense of Vision-Language Models against Malicious Prompts

VLMShield: defense mechanism for vision-language models against malicious prompt attacks using multimodal feature extraction.

Ax Mohammed Nowaz Rabbani Chowdhury, Kaoutar El Maghraoui, Hsinyu Tsai, Naigang Wang, Geoffrey W. Burr, Liu Liu, Meng Wang 21d ago

Efficient Quantization of Mixture-of-Experts with Theoretical Generalization Guarantees

Method for post-training quantization of sparse Mixture-of-Experts models with theoretical generalization guarantees.

Ax Yao Sun, Bo Hu, Jose Principe 21d ago

Time-Series Classification with Multivariate Statistical Dependence Features

Framework for time-series classification using cross density ratio instead of correlation-based statistics.

Ax Bryan Cheng, Jasper Zhang 21d ago

When Does Context Help? A Systematic Study of Target-Conditional Molecular Property Prediction

Systematic study of target context conditioning for molecular property prediction across protein families and data regimes.

Ax Nan Zhang, Zishuo Wang, Shuyu Huang, Georgios Diamantopoulos, Nikos Tziritas, Panagiotis Oikonomou, Georgios Theodoropoulos 21d ago

TwinLoop: Simulation-in-the-Loop Digital Twins for Online Multi-Agent Reinforcement Learning

TwinLoop: simulation-in-the-loop digital twin framework for online multi-agent reinforcement learning with context shifts.

Ax Xiancheng Wang, Lin Wang, Rui Wang, Zhibo Zhang, Minghang Zhao, Xiaoheng Zhang, Zhongyue Tan, Kaitai Mao 21d ago

PD-SOVNet: A Physics-Driven Second-Order Vibration Operator Network for Estimating Wheel Polygonal Roughness from Axle-Box Vibrations

Physics-driven neural network for estimating wheel polygonal roughness from vibration signals in rail vehicles.

Ax Zheng Jiang, Nan He, Yiming Chen, Lifeng Sun 21d ago

SubFLOT: Submodel Extraction for Efficient and Personalized Federated Learning via Optimal Transport

SubFLOT: federated learning method using optimal transport for personalized submodel extraction in heterogeneous settings.

Ax Zhengyang Ai, Zikang Shan, Xiaodong Ai, Jingxian Tang, Hangkai Hu, Pinyan Lu 21d ago

SHAPE: Stage-aware Hierarchical Advantage via Potential Estimation for LLM Reasoning

SHAPE framework for LLM reasoning using stage-aware hierarchical advantage estimation to improve process supervision efficiency.

Ax Devender Singh, Tarun Sheel 21d ago

FlowAdam: Implicit Regularization via Geometry-Aware Soft Momentum Injection

FlowAdam: hybrid optimizer augmenting Adam with geometry-aware soft momentum injection for handling parameter couplings.

Ax Yue Fang, Weibin Liao, Yuxin Guo, Jiaran Gao, Hongxin Ding, Jinyang Zhang, Xinke Jiang, Zhibang Yang, Junfeng Zhao, Yasha Wang, Liantao Ma 21d ago

GraphWalker: Graph-Guided In-Context Learning for Clinical Reasoning on Electronic Health Records

GraphWalker: graph-guided in-context learning framework for LLM-based clinical reasoning on electronic health records.

Ax Qipeng Zhan, Zhuoping Zhou, Li Shen 21d ago

Towards Accurate and Calibrated Classification: Regularizing Cross-Entropy From A Generative Perspective

Method for improving classification calibration using generative perspective to regularize cross-entropy loss in deep networks.

Ax Qipeng Zhan, Zhuoping Zhou, Zexuan Wang, Qi Long, Li Shen 21d ago

Bi-Lipschitz Autoencoder With Injectivity Guarantee

Bi-Lipschitz autoencoder with injectivity guarantee for dimensionality reduction while preserving manifold geometry.

Ax Shengchao Chen, Guodong Long, Dikai Liu, Jing Jiang 21d ago

Bi-level Heterogeneous Learning for Time Series Foundation Models: A Federated Learning Approach

Federated learning approach for training time series foundation models using bi-level heterogeneous learning to address gradient conflicts.

Ax Fumito Kimura, Jun Ohkubo 21d ago

Extraction of linearized models from pre-trained networks via knowledge distillation

Framework for extracting linearized neural network models via knowledge distillation for photonic hardware compatibility.

Ax Zinaid Kapi\'c, Vladimir Ja\'cimovi\'c 21d ago

Busemann energy-based attention for emotion analysis in Poincar\'e discs

EmBolic: hyperbolic deep learning architecture for emotion analysis from text using Busemann energy-based attention.

Ax Robert C. Williamson 21d ago

The Rhetoric of Machine Learning

Philosophical examination of machine learning through rhetoric lens, arguing ML is inherently rhetorical rather than objective.

Ax Marshall Brett 21d ago

Geometric Properties of the Voronoi Tessellation in Latent Semantic Manifolds of Large Language Models

Empirical study of Voronoi tessellations in LLM latent spaces, validating scaling laws of expressibility gaps.

Ax Andrea Pollastro, Andrea Apicella, Francesco Isgr\`o, Roberto Prevete 21d ago

Instance-Adaptive Parametrization for Amortized Variational Inference

Instance-adaptive variational autoencoder addressing amortization gap in latent variable models through per-instance parametrization.

Ax Zhixiong Zhao, Zukang Xu, Zhixuan Chen, Dawei Yang 21d ago

MoBiE: Efficient Inference of Mixture of Binary Experts under Post-Training Quantization

MoBiE: Binarization framework for efficient inference of Mixture-of-Experts LLMs with post-training quantization techniques.

Ax Dihong Jiang, Ruoqi Cao, Zhiyuan Dang, Li Huang, Qingsong Zhang, Zhiyu Wang, Shihao Piao, Shenggao Zhu, Jianlong Chang, Zhouchen Lin, Qi Tian 21d ago

OmniTabBench: Mapping the Empirical Frontiers of GBDTs, Neural Networks, and Foundation Models for Tabular Data at Scale

OmniTabBench: Large-scale benchmark comparing GBDTs, neural networks, and foundation models on tabular data with 100+ datasets.

Ax Minglu Liu, Cunchen Hu, Liangliang Xu, Fengming Tang, Ruijia Wang, Fu Yu 21d ago

STQuant: Spatio-Temporal Adaptive Framework for Optimizer Quantization in Large Multimodal Model Training

STQuant framework for adaptive quantization of optimizer states during large multimodal model training to reduce memory costs.

Ax Hyukjun Yang, Han-Dong Lim, Donghwan Lee 21d ago

Contraction-Aligned Analysis of Soft Bellman Residual Minimization with Weighted Lp-Norm for Markov Decision Problem

Theoretical analysis of Bellman residual minimization for solving Markov decision processes under linear function approximation.

Ax Tianyue Yang, Xiao Xue 21d ago

MENO: MeanFlow-Enhanced Neural Operators for Dynamical Systems

Neural operator enhancement method for dynamical systems combining Fourier-based operators with diffusion-based high-frequency recovery.

Ax Alessandro Pasqui, Jim Martin Catacora Ocana, Anshuman Sinha, Matthieu Perez, Fabrice Delbary, Giorgio Gosti, Mattia Miotto, Domenico Caudo, Maxence Ernoult, Herv\'e Turlier 21d ago

VertAX: a differentiable vertex model for learning epithelial tissue mechanics

JAX-based differentiable framework for vertex-modeling epithelial tissue mechanics with automatic differentiation and GPU acceleration.

Ax Charbel Bou Chaaya, Mehdi Bennis 21d ago

Equivariant Multi-agent Reinforcement Learning for Multimodal Vehicle-to-Infrastructure Systems

Decentralized multi-agent RL approach for vehicle-to-infrastructure systems using equivariant neural networks.

Ax Yitong Li, Junsong Chen, Shuchen Xue, Pengcuo Zeren, Siyuan Fu, Dinghao Yang, Yangyang Tang, Junjie Bai, Ping Luo, Song Han, Enze Xie 21d ago

FP4 Explore, BF16 Train: Diffusion Reinforcement Learning via Efficient Rollout Scaling

Efficient scaling technique for diffusion RL post-training using low-precision exploration and higher-precision training.

Ax Andoni Irazusta Garmendia 21d ago

A First Guess is Rarely the Final Answer: Learning to Search in the Travelling Salesperson Problem

Neural method for learning search policies in Traveling Salesperson Problem, training models to iteratively improve solutions.

Ax Ioannis Kyprakis, Vasileios Skaramagkas, Georgia Karanasiou, Lampros Lakkas, Andri Papakonstantinou, Domen Ribnikar, Kalliopi Keramida, Dorothea Tsekoura, Ketti Mazzocco, Anastasia Constantinidou, Konstantinos Marias, Dimitrios I. Fotiadis, Manolis Tsiknakis 21d ago

Frailty Estimation in Elderly Oncology Patients Using Multimodal Wearable Data and Multi-Instance Learning

Frailty assessment framework for elderly oncology patients using multimodal wearable data and multi-instance learning.