Isolater - Feed

Ax Jongsoo Lee, Jangwon Kim, Soohee Han 4/7/2026

Delayed Homomorphic Reinforcement Learning for Environments with Delayed Feedback

Reinforcement learning approach for handling delayed feedback by replacing state augmentation with homomorphic methods to reduce sample complexity.

Ax Jonathan Katzy, Razvan-Mihai Popescu, Erik Mekkes, Arie van Deursen, Maliheh Izadi 4/7/2026

Automated Attention Pattern Discovery at Scale in Large Language Models

Mechanistic interpretability method for discovering repeated attention patterns in large language models at scale without resource-intensive controlled settings.

Ax Renzo G. Soatto, Anders Hoel, Greycen Ren, Shorna Alam, Stephen Bates, Nikolaos P. Daskalakis, Caroline Uhler, Maria Skoularidou 4/7/2026

CountsDiff: A Diffusion Model on the Natural Numbers for Generation and Imputation of Count-Based Data

CountsDiff: diffusion model framework for generating and imputing count-based discrete ordinal data using survival probability schedules.

Ax Haocheng Ju, Guoxiong Gao, Jiedong Jiang, Bin Wu, Zeming Sun, Leheng Chen, Yutong Wang, Yuefeng Wang, Zichen Wang, Wanyi He, Peihao Wu, Liang Xiao, Ruochuan Liu, Bryan Dai, Bin Dong 4/7/2026

Automated Conjecture Resolution with Formal Verification

Framework for automated mathematical conjecture resolution combining LLMs with formal verification to improve reliability of research-level mathematical problem solving.

Ax Dipkumar Patel 4/7/2026

Representational Collapse in Multi-Agent LLM Committees: Measurement and Diversity-Aware Consensus

Research on representational collapse in multi-agent LLM committees using majority voting, measuring agent diversity via cosine similarity and effective rank on mathematical reasoning tasks.

Ax Jonas De Schouwer, Haitz S\'aez de Oc\'ariz Borde, Xiaowen Dong 4/7/2026

k-Maximum Inner Product Attention for Graph Transformers and the Expressive Power of GraphGPS The Expressive Power of GraphGPS

k-Maximum inner product attention for graph transformers addressing quadratic complexity while maintaining expressive power of GraphGPS.

Ax Giansalvo Cirrincione, Rahul Ranjeev Kumar 4/7/2026

Collapse-Free Prototype Readout Layer for Transformer Encoders

DDCL-Attention: Prototype-based readout layer for transformer encoders using soft probabilistic token matching for compact summaries.

Ax Dharmesh Tailor, Nicol\`o Felicioni, Kamil Ciosek 4/7/2026

A Bayesian Information-Theoretic Approach to Data Attribution

Bayesian information-theoretic approach to training data attribution for tracing model predictions to influential training examples.

Ax Soham Gadgil, Chris Lin, Su-In Lee 4/7/2026

Where to Steer: Input-Dependent Layer Selection for Steering Improves LLM Alignment

Method for input-dependent layer selection in steering vectors to improve LLM alignment at inference time, adapting intervention layer per input.

Ax Xiwen Chen, Jingjing Wang, Wenhui Zhu, Peijie Qiu, Xuanzhao Dong, Hejian Sang, Zhipeng Wang, Alborz Geramifard, Feng Luo 4/7/2026

SODA: Semi On-Policy Black-Box Distillation for Large Language Models

SODA: Semi on-policy knowledge distillation method for LLMs balancing off-policy simplicity with on-policy effectiveness without adversarial training instability.

Ax Yaoze Guo, Shana Moothedath 4/7/2026

Provable Multi-Task Reinforcement Learning: A Representation Learning Framework with Low Rank Rewards

Theoretical research on multi-task representation learning for reinforcement learning with shared representations across related RL tasks with different rewards.

Ax Aniketh Iyengar, Jiaqi Han, Pengwei Sun, Mingjian Jiang, Jianwen Xie, Stefano Ermon 4/7/2026

Align Your Structures: Generating Trajectories with Structure Pretraining for Molecular Dynamics

Framework combining structure pretraining with diffusion models for generating molecular dynamics trajectories with limited MD data.

Ax Hui Sun, Yun-Ji Zhang, Zheng Xie, Ren-Biao Liu, Yali Du, Xin-Ye Li, Ming Li 4/7/2026

ACES: Who Tests the Tests? Leave-One-Out AUC Consistency for Code Generation

ACES: method for selecting LLM-generated code using LLM-generated tests via leave-one-out AUC consistency without determining test correctness.

Ax Yifu Ding, Xinhao Zhang, Jinyang Guo 4/7/2026

Diagonal-Tiled Mixed-Precision Attention for Efficient Low-Bit MXFP Inference

Low-bit mixed-precision attention kernel using MXFP for efficient transformer inference with reduced memory bandwidth.

Ax Yifu Ding, Xianglong Liu, Shenghao Jin, Jinyang Guo, Jiwen Lu 4/7/2026

BWTA: Accurate and Efficient Binarized Transformer by Algorithm-Hardware Co-design

BWTA: binarized transformer quantization scheme with ternary activations and algorithm-hardware co-design for efficient inference.

Ax Shenzhi Yang, Guangcheng Zhu, Bowen Song, Sharon Li, Haobo Wang, Xing Zheng, Yingfan Ma, Zhongqi Chen, Weiqiang Wang, Gang Chen 4/7/2026

Can LLMs Learn to Reason Robustly under Noisy Supervision?

Analysis of LLM reasoning models under noisy labels in reinforcement learning with verifiable rewards, identifying label noise vulnerabilities.

Ax Ozgur Yilmaz 4/7/2026

ArrowFlow: Hierarchical Machine Learning in the Space of Permutations

ArrowFlow: novel ML architecture operating in permutation space using ranking filters and permutation-matrix updates without gradients.

Ax Xuelin Zhang, Hong Chen, Bin Gu, Tieliang Gong, Feng Zheng 4/7/2026

Fine-grained Analysis of Stability and Generalization for Stochastic Bilevel Optimization

Generalization analysis of stochastic bilevel optimization with applications to hyperparameter optimization, meta-learning, and RL.

Ax Milo Coombs 4/7/2026

Spectral Path Regression: Directional Chebyshev Harmonics for Interpretable Tabular Learning

Spectral Path Regression using directional Chebyshev harmonics for interpretable learning on tabular data without exponential scaling.

Ax Prashant C. Raju 4/7/2026

The Geometric Alignment Tax: Tokenization vs. Continuous Geometry in Scientific Foundation Models

Analysis of geometric alignment cost in scientific foundation models for biology/physics, showing discrete tokenization degrades continuous geometry preservation.

Ax Qian Zhou, Yuanyun Zhang, Shi Li 4/7/2026

Uncertainty-Aware Foundation Models for Clinical Data

Framework for uncertainty-aware foundation models on clinical data, addressing incomplete and irregular measurements in healthcare.

Ax Haonian Ji, Kaiwen Xiong, Siwei Han, Peng Xia, Shi Qiu, Yiyang Zhou, Jiaqi Liu, Jinlong Li, Bingzhou Li, Zeyu Zheng, Cihang Xie, Huaxiu Yao 4/7/2026

ClawArena: Benchmarking AI Agents in Evolving Information Environments

ClawArena benchmark for evaluating AI agents in dynamic environments with evolving information, contradictions, and implicit user feedback.

Ax Muhammad Rizwan Awan, Volker Pickert, Muhammad Waqar Ashraf, Saleh Ali, Farshid Mahmouditabar, Shafiq Odhano 4/7/2026

Towards Agentic Defect Reasoning: A Graph-Assisted Retrieval Framework for Laser Powder Bed Fusion

Graph-assisted retrieval framework for reasoning about defects in laser powder bed fusion manufacturing using structured scientific knowledge.

Ax Aniruddh G. Puranic, Sebastian Schirmer, John S. Baras, Calin Belta 4/7/2026

Learning from Imperfect Demonstrations via Temporal Behavior Tree-Guided Trajectory Repair

Framework using Temporal Behavior Trees to repair suboptimal trajectories before using them for robot control policy learning.

Ax Charafeddine Mouzouni 4/7/2026

Three Phases of Expert Routing: How Load Balance Evolves During Mixture-of-Experts Training

Analysis of token routing in Mixture-of-Experts models reveals three-phase training trajectory for load balance evolution.

Ax Yancheng Huang, Changsheng Wang, Chongyu Fan, Yicheng Lang, Bingqi Shang, Yang Zhang, Mingyi Hong, Qing Qu, Alvaro Velasquez, Sijia Liu 4/7/2026

Subspace Control: Turning Constrained Model Steering into Controllable Spectral Optimization

Method for constrained model steering of LLMs addressing safety/privacy requirements via spectral subspace optimization.

Ax Wenhao Chi, \c{S}. \.Ilker Birbil 4/7/2026

Learning An Interpretable Risk Scoring System for Maximizing Decision Net Benefit

Risk scoring system optimizing net benefit using sparse integer linear programming for high-stakes decision-making.

Ax Aobo Chen, Chenxu Zhao, Chenglin Miao, Mengdi Huai 4/7/2026

Towards Unveiling Vulnerabilities of Large Reasoning Models in Machine Unlearning

Study of vulnerabilities in large reasoning models when applying machine unlearning techniques to remove influence of specific data.

Ax Mahmoud Srewa, Tianyu Zhao, Salma Elmalaki 4/7/2026

APPA: Adaptive Preference Pluralistic Alignment for Fair Federated RLHF of LLMs

Federated RLHF method for fair LLM alignment across diverse human preferences without centralizing preference data.

Ax Hristo Petkov, Calum MacLellan, Feng Dong 4/7/2026

DAGAF: A directed acyclic generative adversarial framework for joint structure learning and tabular data synthesis

Dual-step generative framework combining causal structure learning with tabular data synthesis using directed acyclic GANs.

Ax Albert Nasybullin, Semen Kurkin 4/7/2026

Convolutional Neural Network and Adversarial Autoencoder in EEG images classification

Computer vision and neural network approach for classifying human brain activity from EEG data during hand movement.

Ax Albert Nasybullin, Vladimir Maksimenko, Semen Kurkin 4/7/2026

How Long short-term memory artificial neural network, synthetic data, and fine-tuning improve the classification of raw EEG data

Pipeline combining LSTM, synthetic data, and fine-tuning for EEG classification on implicit visual stimuli tasks.

Ax Zequn Chen, Wesley J. Marrero 4/7/2026

Boosted Distributional Reinforcement Learning: Analysis and Healthcare Applications

Analysis of distributional reinforcement learning for complex domains like healthcare, addressing heterogeneous groups under uncertainty.

Ax Xiuyuan Cheng, Yunqin Zhu, Yao Xie 4/7/2026

Generative models for decision-making under distributional shift

Tutorial on using flow- and score-based generative models for decision-making under distributional shift in operations research.

Ax Karthik Suresh, Amine Ben Khalifa, Li Zhang, Wei-ting Hsu, Fangzheng Wu, Vinay More, Asim Kadav 4/7/2026

CPT: Controllable and Editable Design Variations with Language Models

System for generating editable design variations using decoder-only language model with Creative Markup Language representation.

Ax Narim Jeong, Donghwan Lee 4/7/2026

Finite-Time Analysis of Q-Value Iteration for General-Sum Stackelberg Games

Theoretical analysis of Q-value iteration convergence in multi-agent Stackelberg games using control-theoretic perspective.

Ax Hiroshi Takahashi, Tomoharu Iwata, Atsutoshi Kumagai, Sekitoshi Kanai, Masanori Yamada, Kosuke Nishida, Kazutoshi Shinoda 4/7/2026

Relative Density Ratio Optimization for Stable and Statistically Consistent Model Alignment

Research on aligning LLMs with human preferences using relative density ratio optimization without assuming specific preference models, improving statistical consistency.

Ax Seoyoung Park, Haemin Lee, Hankook Lee 4/7/2026

Is Prompt Selection Necessary for Task-Free Online Continual Learning?

Study questioning necessity of prompt selection in task-free online continual learning for non-stationary data streams.

Ax Henrik Krauss, Takehisa Yairi 4/7/2026

Estimating Central, Peripheral, and Temporal Visual Contributions to Human Decision Making in Atari Games

Ablation framework to estimate contributions of central, peripheral, and temporal visual information to human decision-making in Atari games.

Ax Fatemeh Khadem, Sajad Mousavi, Yi Fang, Yuhong Liu 4/7/2026

DP-OPD: Differentially Private On-Policy Distillation for Language Models

DP-OPD: differentially private on-policy distillation method for compressing LLMs on sensitive data while maintaining privacy guarantees.

Ax Zhe Feng, Shilong Tao, Haonan Sun, Shaohan Chen, Zhanxing Zhu, Yunhuai Liu 4/7/2026

MAVEN: A Mesh-Aware Volumetric Encoding Network for Simulating 3D Flexible Deformation

MAVEN: mesh-aware volumetric encoding network for simulating 3D flexible deformation using graph neural networks on mesh structures.

Ax Liwei Deng, Qingxiang Liu, Xinhe Niu, Shengchao Chen, Sheng Sun, Yuankai Wu, Guodong Long, Yuxuan Liang 4/7/2026

Discrete Prototypical Memories for Federated Time Series Foundation Models

Discrete Prototypical Memories approach for federated time series foundation models using LLMs while preserving data privacy.

Ax Tauhid Khan 4/7/2026

Isokinetic Flow Matching for Pathwise Straightening of Generative Flows

Isokinetic Flow Matching introduces pathwise acceleration regularization to improve few-step sampling in flow-based generative models.

Ax Ziwei Li, Yuang Ma, Yi Kang 4/7/2026

SLaB: Sparse-Lowrank-Binary Decomposition for Efficient Large Language Models

SLaB: sparse-lowrank-binary decomposition framework for efficient LLM compression maintaining performance at high compression ratios.

Ax Qiang He, Yucheng Yang, Tianyi Zhou, Meng Fang, Mykola Pechenizkiy, Setareh Maghsudi 4/7/2026

One Model for All: Multi-Objective Controllable Language Models

Multi-objective controllable language models framework enabling personalized alignment with varying human preferences beyond fixed reward optimization.

Ax Hengshuai Yao, Xing Chen, Ahmed Murtadha, Guan Wang 4/7/2026

GAIN: Multiplicative Modulation for Domain Adaptation

GAIN: multiplicative modulation technique for domain adaptation in LLMs, preventing catastrophic forgetting through feature re-emphasis.

Ax Ole Delzer, Sidney Bender 4/7/2026

Reproducibility study on how to find Spurious Correlations, Shortcut Learning, Clever Hans or Group-Distributional non-robustness and how to fix them

Reproducibility study on spurious correlations and shortcut learning in DNNs, comparing frameworks for ensuring models use causally relevant features.

Ax Mark Braverman, Roi Livni, Yishay Mansour, Shay Moran, Kobbi Nissim 4/7/2026