Isolater - Feed

Ax Anbang Ruan, Xing Zhang 4/9/2026

AgentCity: Constitutional Governance for Autonomous Agent Economies via Separation of Power

Framework for governing autonomous AI agent economies through constitutional separation of powers architecture.

Ax Philipp D. Siedler 4/9/2026

Strategic Persuasion with Trait-Conditioned Multi-Agent Systems for Iterative Legal Argumentation

Multi-agent LLM simulation framework for legal argumentation with trait-conditioned agents in adversarial game-theoretic setting.

Ax Mehdi Hosseinzadeh, King Hang Wong, Feras Dayoub 4/9/2026

KITE: Keyframe-Indexed Tokenized Evidence for VLM-Based Robot Failure Analysis

Training-free method using vision-language models to analyze robot execution failures via keyframe tokenization.

Ax Xue Qin, Simin Luan, Cong Yang, Zhijun Li 4/9/2026

AEROS: A Single-Agent Operating Architecture with Embodied Capability Modules

Single-agent robotic architecture with modular capabilities for unified intelligence organization and execution control.

Ax Minh Tam Pham, Trinh Pham, Tong Chen, Hongzhi Yin, Quoc Viet Hung Nguyen, Thanh Tam Nguyen 4/9/2026

AV-SQL: Decomposing Complex Text-to-SQL Queries with Agentic Views

Agentic approach using LLM agents to decompose complex text-to-SQL queries into simpler steps for multi-table reasoning.

Ax Hongru Ji, Yuyin Fan, Meng Zhao, Xianghua Li, Lianwei Wu, Chao Gao 4/9/2026

STRIDE-ED: A Strategy-Grounded Stepwise Reasoning Framework for Empathetic Dialogue Systems

Framework for empathetic dialogue systems using strategy-aware multi-stage reasoning and LLM-based response generation.

Ax Yongchao Wu, Aron Henriksson 4/9/2026

The Impact of Steering Large Language Models with Persona Vectors in Educational Applications

Study on persona vector steering for personalizing LLM outputs in educational settings; shows reduced answer quality on open-ended tasks.

Ax Radu Negulescu 4/9/2026

Information as Structural Alignment: A Dynamical Theory of Continual Learning

Informational Buildup Framework addressing catastrophic forgetting in continual learning through alternative parameter storage mechanisms.

Ax Haichang Li, Qinshi Zhang, Piaohong Wang, Zhicong Lu 4/9/2026

Mixed-Initiative Context: Structuring and Managing Context for Human-AI Collaboration

Framework for dynamically organizing and managing context in multi-turn human-AI collaboration workflows with hierarchical structure awareness.

Ax Qichen Wang, Keyu Li, Ozan Alp Topal, \"Ozlem Tugfe Demir, Mustafa Ozger, Cicek Cavdar 4/9/2026

Energy Saving for Cell-Free Massive MIMO Networks: A Multi-Agent Deep Reinforcement Learning Approach

Multi-agent deep reinforcement learning algorithm for energy optimization in cell-free massive MIMO networks.

Ax Ryan Lingo, Rajeev Chhajer 4/9/2026

Dynamic Context Evolution for Scalable Synthetic Data Generation

Framework addressing cross-batch mode collapse in LLM synthetic data generation through Dynamic Context Evolution.

Ax Nathan Lambert, Florian Brand 4/9/2026

The ATOM Report: Measuring the Open Language Model Ecosystem

Comprehensive ecosystem analysis of ~1.5K open language models, documenting adoption trends and builders of leading models.

Ax Yen-Shan Chen, Sian-Yao Huang, Cheng-Lin Yang, Yun-Nung Chen 4/9/2026

TraceSafe: A Systematic Assessment of LLM Guardrails on Multi-Step Tool-Calling Trajectories

Benchmark assessing safety guardrails for LLMs in multi-step tool-calling agent trajectories, introducing TraceSafe-Bench.

Ax Hamayoon Behmanush, Freshta Akhtari, Ingmar Weber, Vikram Kamath Cannanure 4/9/2026

Designing Safe and Accountable GenAI as a Learning Companion with Women Banned from Formal Education

Research on designing safe and accountable generative AI as learning companion for women in surveillance-restricted contexts.

Ax Yuanhang Li 4/9/2026

Validated Intent Compilation for Constrained Routing in LEO Mega-Constellations

System translating natural language operator intents into routing constraints for LEO satellite networks using GNN and LLM components.

Ax Nusrat Sultana, Abdullah Muhammad Moosa, Kazi Afzalur Rahman, Sajal Chandra Banik 4/9/2026

A Systematic Study of Retrieval Pipeline Design for Retrieval-Augmented Medical Question Answering

Systematic study analyzing retrieval pipeline components for retrieval-augmented generation in medical question answering systems.

Ax Luca Pennati, Andong Hu, Ivy Peng, Lukas M\"ullender, Stefano Markidis 4/9/2026

Making Room for AI: Multi-GPU Molecular Dynamics with Deep Potentials in GROMACS

Integration of DeePMD-kit neural network potentials into GROMACS for GPU-accelerated molecular dynamics simulations.

Ax Guo Gan, Yuxuan Ding, Cong Chen, Yuwei Ren, Yin Huang, Hong Zhou 4/9/2026

Android Coach: Improve Online Agentic Training Efficiency with Single State Multiple Actions

Framework improving online reinforcement learning efficiency for Android agents by enabling multiple actions per state.

Ax Eduard Frankford, Erik Cikalleshi, Ruth Breu 4/9/2026

Chatbot-Based Assessment of Code Understanding in Automated Programming Assessment Systems

Research on using conversational LLMs for automated programming assessment to evaluate student code understanding beyond functional correctness.

Ax Jackson Petty, Jaulie Goe, Tal Linzen 4/9/2026

Evaluating In-Context Translation with Synchronous Context-Free Grammar Transduction

Study of LLMs' in-context learning for machine translation using grammatical descriptions, focusing on low-resource languages.

Ax Priscilla Kyei Danso, Mohammad Saqib Hasan, Niranjan Balasubramanian, Omar Chowdhury 4/9/2026

Syntax Is Easy, Semantics Is Hard: Evaluating LLMs for LTL Translation

Research evaluating LLMs' ability to translate natural language into Linear Temporal Logic formulas for security and privacy specifications.

Ax Pascal J. Sager, Benjamin Meyer, Peng Yan, Rebekka von Wartburg-Kottler, Layan Etaiwi, Aref Enayati, Gabriel Nobel, Ahmed Abdulkadir, Benjamin F. Grewe, Thilo Stadelmann 4/9/2026

A Comprehensive Survey of Agents for Computer Use: Foundations, Challenges, and Future Directions

Survey of agents for computer use (ACUs) covering systems that execute complex tasks on digital devices via natural language instructions and low-level actions.

Ax Aabid Karim, Abdul Karim, Bhoomika Lohana, Matt Keon, Jaswinder Singh, Abdul Sattar 4/9/2026

Lost in Cultural Translation: Do LLMs Struggle with Math Across Cultural Contexts?

Analysis of 14 LLM models showing mathematical reasoning performance drops 0.3-5.9% when problems use unfamiliar cultural contexts despite identical logic.

Ax Timoth\'ee Loranchet, Charles K. Assaad 4/9/2026

Local Markov Equivalence for PC-style Local Causal Discovery and Identification of Controlled Direct Effects

PC algorithm extension for identifying controlled direct effects in causal discovery from Markov equivalence classes using essential graphs.

Ax Xuanle Zhao, Zilin Sang, Yuxuan Li, Qi Shi, Weilun Zhao, Shuo Wang, Duzhen Zhang, Xu Han, Zhiyuan Liu, Maosong Sun 4/9/2026

AutoReproduce: Automatic AI Experiment Reproduction with Paper Lineage

AutoReproduce: multi-agent system for automatic AI experiment reproduction mining implicit knowledge from cited literature through paper lineage.

Ax Yazhou Zhang, Chunwang Zou, Bo Wang, Jing Qin, Prayag Tiwari 4/9/2026

Commander-GPT: Dividing and Routing for Multimodal Sarcasm Detection

Commander-GPT: modular routing framework orchestrating multiple LLMs for multimodal sarcasm detection using military command theory.

Ax Shuquan Lian, Yuhang Wu, Jia Ma, Yifan Ding, Zihan Song, Bingqi Chen, Xiawu Zheng, Hui Li, Rongrong Ji 4/9/2026

UI-AGILE: Advancing GUI Agents with Effective Reinforcement Learning and Precise Inference-Time Grounding

UI-AGILE: GUI agent framework enhancing multimodal LLM agents with effective RL training and inference-time grounding strategies.

Ax Alberto Pozanco, Marianela Morales, Daniel Borrajo, Manuela Veloso 4/9/2026

Planning with Minimal Disruption

Planning algorithms minimizing disruption to initial state while achieving goals through action cost and disruption joint optimization.

Ax Jackson Eshbaugh, Chetan Tiwari, Jorge Silveyra 4/9/2026

Synthetic Homes: A Multimodal Generative AI Pipeline for Residential Building Data Generation under Data Scarcity

Multimodal generative AI pipeline for synthetic residential building data generation addressing data scarcity in energy modeling.

Ax Zaid Khan, Archiki Prasad, Elias Stengel-Eskin, Jaemin Cho, Mohit Bansal 4/9/2026

One Life to Learn: Inferring Symbolic World Models for Stochastic Environments from Unguided Exploration

Learning symbolic world models from unguided single-episode exploration in complex stochastic environments without human guidance or interaction data.

Ax Junxian Li, Xinyue Xu, Sai Ma, Di Zhang, Sichao Li 4/9/2026

Faithful-First Reasoning, Planning, and Acting for Multimodal LLMs

Faithful-First RPA framework for multimodal LLMs using step-wise faithfulness supervision to align reasoning and actions with visual evidence.

Ax Jua Han, Jaeyoon Seo, Jungbin Min, Sieun Choi, Huichan Seo, Jihie Kim, Jean Oh 4/9/2026

Before We Trust Them: Decision-Making Failures in Navigation of Foundation Models

Evaluation of foundation models on spatial reasoning tasks revealing limitations in decision-making not captured by standard navigation metrics.

Ax Edward Y. Chang 4/9/2026

Diagnosing and Mitigating Sycophancy and Skepticism in LLM Causal Judgment

CausalT3 benchmark diagnosing control failures in LLM causal reasoning where models abandon sound reasoning under social pressure or authoritative hints.

Ax Mayank Sharma, Roy Pea, Hari Subramonyam 4/9/2026

ConvoLearn: A Dataset for Fine-Tuning Dialogic AI Tutors

ConvoLearn: dataset of 2,134 tutor-student dialogues for fine-tuning LLM-based AI tutors aligned with dialogic knowledge-building pedagogy.

Ax Tongxi Wang 4/9/2026

FBS: Modeling Native Parallel Reading inside a Transformer

Fovea-Block-Skip Transformer enabling parallel chunk-aware reading in LLMs through causal trainable loops mimicking human skimming and content-adaptive foresight.

Ax Shuo Lu, Jianjie Cheng, Yinuo Xu, Yongcan Yu, Lijun Sheng, Peijie Wang, Siru Jiang, Yongguan Hu, Run Ling, Yihua Shao, Ao Ma, Wei Feng, Lingxiao He, Meng Wang, Qianlong Xie, Xingxing Wang, Nicu Sebe, Ran He, Jian Liang 4/9/2026

Do MLLMs Really Understand Space? A Mathematical Reasoning Evaluation

Evaluation of multimodal LLM spatial reasoning capabilities showing models significantly underperform humans on mathematical spatial reasoning tasks.

Ax Xin An, Jingyi Cai, Xiangyang Chen, Huayao Liu, Peiting Liu, Peng Wang, Bei Yang, Xiuwen Zhu, Yongfan Chen, Yan Gao, Yuan Gao, Baoyu Hou, Guangzheng Hu, Shuzhao Li, Weixu Qiao, Weidong Ren, Yanan Wang, Boyu Yang, Fan Yang, Jiangtao Zhang, Lixin Zhang, Lin Qu, Hu Wei, Xiaoxiao Xu, Bing Zhao 4/9/2026

Logics-Parsing-Omni Technical Report

Omni Parsing framework with unified taxonomy for multimodal parsing across documents, images, and audio-visual streams using hierarchical perception and cognition levels.

Ax Tianhao Qian, Zhuoxuan Li, Jinde Cao, Xinli Shi, Leszek Rutkowski 4/9/2026

Resource-constrained Amazons chess decision framework integrating large language models and graph attention

Lightweight hybrid framework combining LLMs and graph attention for Amazons chess in resource-constrained environments without extensive datasets.

Ax Pablo de los Riscos, Fernando J. Corbacho, Michael A. Arbib 4/9/2026

Working Paper: Towards a Category-theoretic Comparative Framework for Artificial General Intelligence

Category-theoretic framework for comparing and analyzing AGI definitions and benchmarks using algebraic approaches.

Ax Zekai Ye, Qiming Li, Xiaocheng Feng, Ruihan Chen, Ziming Li, Haoyu Ren, Kun Chen, Dandan Tu, Bing Qin 4/9/2026

Not All Tokens See Equally: Perception-Grounded Policy Optimization for Large Vision-Language Models

Token-level perception-grounded policy optimization for vision-language models addressing diluted learning signals in multimodal reasoning through selective token weighting.

Ax Yu Li, Haoyu Luo, Yuejin Xie, Yuqian Fu, Zhonghao Yang, Shuai Shao, Qihan Ren, Wanying Qu, Yanwei Fu, Yujiu Yang, Jing Shao, Xia Hu, Dongrui Liu 4/9/2026

ATBench: A Diverse and Realistic Agent Trajectory Benchmark for Safety Evaluation and Diagnosis

ATBench: trajectory-level safety benchmark for LLM-based agents evaluating multi-step interactions with structured diversity and realistic long-horizon interactions.

Ax Chao Li, Yuru Wang, Chunyi Zhao 4/9/2026

Domain-Contextualized Inference: A Computable Graph Architecture for Explicit-Domain Reasoning

Computation-substrate-agnostic inference architecture with domain as explicit parameter, enabling domain-scoped pruning and transparent inference chains across symbolic, neural, vector, and hybrid substrates.

Ax Xiangyi Li, Kyoung Whan Choe, Yimin Liu, Xiaokun Chen, Chujun Tao, Bingran You, Wenbo Chen, Zonglin Di, Jiankai Sun, Shenghan Zheng, Jiajun Bao, Yuanli Wang, Weixiang Yan, Yiyuan Li, Han-chung Lee 4/9/2026

ClawsBench: Evaluating Capability and Safety of LLM Productivity Agents in Simulated Workspaces

ClawsBench is a benchmark for evaluating LLM agents on realistic productivity tasks like email and scheduling in simulated multi-service stateful environments.

Ax Yitong Zhu, Yuxuan Jiang, Guanxuan Jiang, Bojing Hou, Peng Yuan Zhou, Ge Lin Kan, Yuyang Wang 4/9/2026

QA-MoE: Towards a Continuous Reliability Spectrum with Quality-Aware Mixture of Experts for Robust Multimodal Sentiment Analysis

QA-MoE proposes a Mixture of Experts approach for multimodal sentiment analysis that adapts to varying input reliability and modality missingness in real-world scenarios.

Ax Eduardo C. Garrido-Merch\'an, \'Alvaro L\'opez L\'opez 4/9/2026

An Automated Survey of Generative Artificial Intelligence: Large Language Models, Architectures, Protocols, and Applications

Comprehensive survey of generative AI covering LLM architectures, deployment protocols, and applications as of early 2026.

Ax Ximing Xing, Chuang Wang, Haitao Zhou, Jing Zhang, Qian Yu, Dong Xu 4/9/2026