Ax Yuxi Cai, Lan Li, Feiqing Huang, Guodong Li 5/5/2026

ParaRNN: An Interpretable and Parallelizable Recurrent Neural Network for Time-Dependent Data

ParaRNN combines interpretability with parallelizability for time-series modeling as nonlinear extension of ARMA models.

Ax Sergio Orozco, Tushar Kusnur, Brandon May, George Konidaris, Laura Herlant 5/5/2026

Learning Equivariant Neural-Augmented Object Dynamics From Few Interactions

PIEGraph combines neural networks with physics constraints for data-efficient learning of deformable object dynamics in robotic manipulation.

Ax Davide Nobile, Philipp Grohs 5/5/2026

Robust and Fast Training via Per-Sample Clipping

Per-sample gradient clipping method for robust SGD with optimal convergence under heavy-tailed gradient noise and high-probability guarantees.

Ax Anahita Golrang, Kshitij Sharma, olga viberg 5/5/2026

ProPACT: A Proactive AI-Driven Adaptive Collaborative Tutor for Pair Programming

ProPACT adaptive tutor uses multimodal dyadic learning models with joint visual attention and mental effort for collaborative pair programming.

Ax Christine P Lee, Xinyu Jessica Wang, Aws Albarghouthi, David Porfirio, Bilge Mutlu 5/5/2026

U-Define: Designing User Workflows for Hard and Soft Constraints in LLM-Based Planning

U-Define system enables users to apply hard and soft constraints in LLM-based planning with improved control and adaptability.

Ax Mohamad Khajezade, Fatemeh H. Fard, Mohamed Sami Shehata 5/5/2026

Standing on the Shoulders of Giants: Stabilized Knowledge Distillation for Cross--Language Code Clone Detection

Knowledge distillation approach for cross-language code clone detection using stabilized training of compact open-source models instead of LLM black-boxes.

Ax Caleb Princewill Nwokocha 5/5/2026

Rule Extraction in Machine Learning: Chat Incremental Pattern Constructor

ChatIPC system extracts interpretable rules from text using incremental symbolic learning with token-transition patterns and similarity-based response construction.

Ax Tian Xu, Ziniu Li, Yang Yu, Zhi-Quan Luo 5/5/2026

Understanding Adversarial Imitation Learning in Small Sample Regime: A Stage-coupled Analysis

Analyzes why adversarial imitation learning performs well with minimal expert data and provides theoretical understanding of the phenomenon.

Ax Amir Leshem 5/5/2026

Near-Optimal Privacy-Preserving Learning for Max-Min Fair Multi-Agent Bandits

Studies fair multi-agent bandit learning with privacy constraints where agents learn optimal allocations without explicit communication.

Ax Hanbeot Park, Yunjeong Cho, Hunhee Kim 5/5/2026

CCNETS: A Modular Causal Learning Framework for Pattern Recognition in Imbalanced Datasets

CCNETS framework addresses class imbalance in ML by integrating data synthesis with classification using causal cooperative networks for improved pattern recognition.

Ax Davide Corsi, Guy Amir, Guy Katz, Alessandro Farinelli 5/5/2026

Analyzing Adversarial Inputs in Deep Reinforcement Learning

Analysis of adversarial input vulnerability in deep reinforcement learning models and reliability concerns for real-world DRL applications.

Ax Alessio Benavoli, Dario Azzimonti 5/5/2026

A tutorial on learning from preferences and choices with Gaussian Processes

Tutorial on preference learning and choice modeling using Gaussian processes for personalized applications across economics and decision theory.

Ax Christian Intern\`o, Elena Raponi, Markus Olhofer, Ali Raza, Thomas B\"ack, Niki van Stein, Yaochu Jin, Barbara Hammer 5/5/2026

Pruning Federated Models through Loss Landscape Analysis and Client Agreement Scoring

AutoFLIP framework prunes federated learning models using loss landscape analysis and client agreement scoring on non-IID heterogeneous data.

Ax Yuxuan Wang, Haixu Wu, Jiaxiang Dong, Yong Liu, Chen Wang, Mingsheng Long, Jianmin Wang 5/5/2026

Deep Time Series Models: A Comprehensive Survey and Benchmark

Comprehensive survey and benchmark of deep time series models covering nonlinear patterns, trends, and recent breakthroughs in time series analysis.

Ax Gleb Bazhenov, Oleg Platonov, Liudmila Prokhorenkova 5/5/2026

GraphLand: Evaluating Graph Machine Learning Models on Diverse Industrial Data

GraphLand benchmark evaluates graph neural networks on diverse industrial datasets beyond academic citation networks for graph foundation model development.

Ax Alejandro Mata Ali, Aitor Moreno Fdez. de Leceta, Jorge L\'opez Rubio 5/5/2026

Anomaly Detection from a Tensor Train Perspective

Tensor network algorithms for anomaly detection using tensor train compression to preserve normal data structure and delete anomalies.

Ax Xavier Warin 5/5/2026

P1-KAN: an effective Kolmogorov-Arnold network with application to hydraulic valley optimization

Kolmogorov-Arnold network variant with error bounds and universal approximation theorems, applied to hydraulic valley optimization.

Ax Samuel J. Bell, Skyler Wang 5/5/2026

The Pragmatic Frames of Spurious Correlations in Machine Learning: Interpreting How and Why They Matter

Analysis framework for understanding spurious correlations in ML systems and their pragmatic implications for model fairness and robustness.

Ax Giacomo Lastrucci, Artur M. Schweidtmann 5/5/2026

ENFORCE: Nonlinear Constrained Learning with Adaptive-depth Neural Projection

ENFORCE architecture enforces nonlinear constraints in neural networks through adaptive projection for safety-critical and domain-specific tasks.

Ax Simone Papicchio, Simone Rossi, Luca Cagliero, Paolo Papotti 5/5/2026

Think2SQL: Reinforce LLM Reasoning Capabilities for Text2SQL

Think2SQL improves LLM reasoning for text-to-SQL translation using reinforcement learning with verifiable rewards for complex multi-table queries.

Ax Maksim Bobrin, Ilya Zisman, Alexander Nikulin, Vladislav Kurenkov, Dmitry Dylov 5/5/2026

Zero-Shot Adaptation of Behavioral Foundation Models to Unseen Dynamics

Behavioral foundation models adapted zero-shot to unseen dynamics using successor measures learned from offline task-agnostic data.

Ax Tajamul Ashraf, Mohammed Mohsen Peerzada, Moloud Abdar, Yutong Xie, Yuyin Zhou, Xiaofeng Liu, Iqra Altaf Gillani, Janibul Bashir 5/5/2026

ATR-Bench: A Federated Learning Benchmark for Adaptation, Trust, and Reasoning

ATR-Bench: unified federated learning benchmark for evaluating adaptation, trust, and reasoning across decentralized collaborative training scenarios.

Ax Sayed Mohammad Vakilzadeh Hatefi, Maximilian Dreyer, Reduan Achtibat, Patrick Kahardipraja, Thomas Wiegand, Wojciech Samek, Alexander Binder, Sebastian Lapuschkin 5/5/2026

Attribution-Guided Pruning for Insight and Control: Circuit Discovery and Targeted Correction in Small-scale LLMs

Mechanistic interpretability research using attribution-guided pruning to discover and correct circuits in small-scale LLMs for improved behavior control.

Ax Andreas Bjerregaard, S{\o}ren Hauberg, Anders Krogh 5/5/2026

Riemannian Generative Decoder

Riemannian generative decoder for learning non-Euclidean data representations without brittle encoder-based density estimation.

Ax A. Bochkov 5/5/2026

Growing Transformers: Modular Composition and Layer-wise Expansion on a Frozen Substrate

Study on training decoder-only Transformers with frozen blocks and fixed token interface, enabling depth expansion with constant active parameters using LoRA.

Ax Sung-Hyun Kim, Geum-Hwan Hwang, In-Chang Baek, Seo-Young Lee, Kyung-Joong Kim 5/5/2026

Multi-Objective Instruction-Aware Representation Learning in Procedural Content Generation RL

Multi-objective instruction-aware learning for procedural content generation using natural language control with RL.

Ax Bokeng Zheng, Jianqiang Zhong, Jiayi Liu, Lei Xue, Xu Chen, Xiaoxi Zhang 5/5/2026

Decentralized Rank Scheduling for Energy-Constrained Multi-Task Federated Fine-Tuning in Edge-Assisted IoV Networks

Decentralized scheduling framework for federated fine-tuning of foundation models across IoV networks with energy constraints.

Ax Cheng Jing, Uvini Balasuriya Mudiyanselage, Woojin Cho, Minju Jo, Anthony Gruber, Kookjin Lee 5/5/2026

Meta-learning Structure-Preserving Dynamics

Meta-learning approach for structure-preserving dynamics discovery applicable across system configurations without retraining.

Ax Rishabh Dixit, Yuan Hui, Rayan Saab 5/5/2026

The Measure of Deception: An Analysis of Data Forging in Machine Unlearning

Analysis of gradient forging attacks in machine unlearning to verify whether models actually forget designated data.

Ax Andrei Chertkov, Artem Basharin, Mikhail Saygin, Evgeny Frolov, Stanislav Straupe, Ivan Oseledets 5/5/2026

Low-rank surrogate modeling and stochastic zero-order optimization for training of neural networks with black-box layers

Surrogate modeling and zero-order optimization for training neural networks with non-differentiable black-box layers.

Ax Lo\"ic Cabannes, Maximilian Beck, Gergely Szilvasy, Matthijs Douze, Maria Lomeli, Jade Copet, Pierre-Emmanuel Mazar\'e, Gabriel Synnaeve, Herv\'e J\'egou 5/5/2026

Short window attention enables long-term memorization

Analysis of sliding window and global attention interaction showing short window length enables effective long-term memorization.

Ax Tavor Z. Baharav, Spyros Dragazis, Aldo Pacchiano 5/5/2026

The Good, the Bad, and the Sampled: a No-Regret Approach to Safe Online Classification

Sequential testing method for binary disease classification with unknown logistic model minimizing costly tests under safety constraints.

Ax Yuxuan Wang, Haixu Wu, Yuezhou Ma, Yuchen Fang, Ziyi Zhang, Yong Liu, Shiyu Wang, Zhou Ye, Yang Xiang, Jianmin Wang, Mingsheng Long 5/5/2026

Exploring Accuracy Law for Deep Time Series Forecasters: An Empirical Study

Empirical study investigating accuracy lower bounds and performance estimation for deep time series forecasting models.

Ax Juan Sebastian Rojas, Chi-Guhn Lee 5/5/2026

Ergodic Risk Measures: Towards a Risk-Aware Foundation for Continual Reinforcement Learning

Risk-aware foundation for continual RL using ergodic risk measures to balance retention and adaptation in lifelong learning.

Ax Yuzheng Hu, Ryan McKenna, Da Yu, Shanshan Wu, Han Zhao, Zheng Xu, Peter Kairouz 5/5/2026

ACTG-ARL: Differentially Private Conditional Text Generation with RL-Boosted Control

Differentially private conditional text generation using RL-boosted control to synthesize datasets while preserving privacy.

Ax Xun Su, Hiroyuki Kasai 5/5/2026

Noise is All You Need: Solving Linear Inverse Problems by Noise Combination Sampling with Diffusion Models

Diffusion model approach for solving linear inverse problems using noise combination sampling without explicit constraint integration.

Ax Yuxiang Chen, Yifan Liu, Xiaoming Xu, Pengle Zhang, Michael Beyer, Martin Rapp, Jun Zhu, Jianfei Chen 5/5/2026

TetraJet-v2: Accurate NVFP4 Training for Large Language Models with Oscillation Suppression and Outlier Control

TetraJet-v2: end-to-end 4-bit fully-quantized training method for LLMs using NVFP4 format with oscillation suppression.

Ax Lipeng Zu, Yu Qian, Shayok Chakraborty, Xiaonan Zhang 5/5/2026

From Static Constraints to Dynamic Adaptation: Sample-Level Constraint Release for Offline-to-Online Reinforcement Learning

Distribution-aware framework (DARE) for offline-to-online reinforcement learning balancing conservatism with online adaptation.

Ax Artyom Sorokin, Nazar Buzun, Alexander Anokhin, Oleg Inozemcev, Egor Vedernikov, Petr Anokhin, Mikhail Burtsev, Trushkov Alexey, Yin Wenshuai, Evgeny Burnaev 5/5/2026

Q-RAG: Long Context Multi-step Retrieval via Value-based Embedder Training

Multi-step retrieval method for RAG systems using value-based embedder training to handle complex questions requiring iterative search.

Ax Dake Bu, Wei Huang, Andi Han, Atsushi Nitanda, Hau-San Wong, Qingfu Zhang, Taiji Suzuki 5/5/2026

Provable Benefit of Curriculum in Transformer Tree-Reasoning Post-Training

Theoretical framework proving curriculum post-training in LLMs yields exponential sample complexity improvements for reasoning tasks.

Ax Zhenchao Tang, Fang Wang, Haohuai He, Jiale Zhou, Tianxu Lv, Jun Zhu, Shouzhi Chen, Minghao Yang, Yu Wang, Jiayang Wu, Yidong Song, Yaokun Li, Jiehui Huang, Bing He, Jianhua Yao 5/5/2026

Aligning LLMs with Biomedical Knowledge using Balanced Fine-Tuning

Balanced fine-tuning approach aligning LLMs with biomedical knowledge by addressing unique uncertainty structures in dense causal chains and rare entities.

Ax Johannes Forkel, Constantin Ruhdorfer, Andreas Bulling, Jakob Foerster 5/5/2026

High entropy leads to symmetry equivariant policies in Dec-POMDPs

Theoretical proof that high entropy regularization in decentralized POMDPs ensures policy convergence to symmetric equivariant solutions.

Ax Soumyadeep Chandra, Sayeed Shafayet Chowdhury, Kaushik Roy 5/5/2026

2D-ThermAl: Physics-Informed Framework for Thermal Analysis of Circuits using Generative AI

Physics-informed generative AI framework for rapid thermal analysis of circuits during early-stage design, accelerating iteration cycles.

Ax Dongseok Kim, Hyoungsun Choi, Mohamed Jismy Aashik Rasool, Gisung Oh 5/5/2026

How Prompts Move Language Model Behavior: Frames, Salience, and Construal as Semantic Control

Cognitive-semantic framework formalizing prompt engineering as semantic control through frame activation, salience manipulation, and task construal.

Ax Yuhan Tang, Kangxin Cui, Jung Ho Park, Yibo Zhao, Xuan Jiang, Haoze He, Jiangbo Yu, Haris Koutsopoulos, Jinhua Zhao 5/5/2026

RAST-MoE-RL: A Regime-Aware Spatio-Temporal MoE Framework for Deep Reinforcement Learning in Ride-Hailing

Regime-aware spatio-temporal mixture of experts with RL for adaptive delayed matching in ride-hailing under dynamic demand conditions.

Ax Susmit Das 5/5/2026

TIME: Temporally Intelligent Meta-reasoning Engine for Context-Triggered Explicit Reasoning

Framework enabling context-triggered explicit reasoning in LLMs, allowing dynamic re-triggering during generation rather than front-loaded thinking tokens.

Ax Nickil Maveli, Antonio Vergari, Shay B. Cohen 5/5/2026

Can LLMs Compress (and Decompress)? Evaluating Code Understanding and Execution via Invertibility

RoundTripCodeEval benchmark evaluating code-LLMs on round-trip consistency through lossless compression algorithm understanding and execution.

Ax Satwik Bhattamishra, Kulin Shah, Michael Hahn, Varun Kanade 5/5/2026

Provably Learning Attention with Queries

Theoretical study of learning Transformer attention mechanisms with black-box query access, proving learnability of single-head attention regressors.

Ax Fatema Siddika, Md Anwar Hossen, Tanwi Mallick, Ali Jannesari 5/5/2026

Split-on-Share: Mixture of Sparse Experts for Task-Agnostic Continual Learning

Mixture of Sparse Experts framework addressing catastrophic forgetting in continual LLM learning by distinguishing task-specific and shared parameters.

Ax Haonan Zhang, Dongxia Wang, Yi Liu, Kexin Chen, Wenhai Wang 5/5/2026

LLM-VA: Resolving the Jailbreak-Overrefusal Trade-off via Vector Alignment

Vector alignment method resolving jailbreak-overrefusal trade-off in LLMs by disentangling answer vector and safety judgment encoding.