Isolater - Feed

Ax Abed K. Musaffar, Ambuj Singh, Francesco Bullo 3/24/2026

Detection of adversarial intent in Human-AI teams using LLMs

Detection of adversarial attacks on LLM-based AI agents in human-AI teams including prompt injection and data poisoning.

Ax Valentina Kuskova, Dmitry Zaytsev, Michael Coppedge 3/24/2026

From Causal Discovery to Dynamic Causal Inference in Neural Time Series

Dynamic Causal Network Autoregression method for discovering causal structure in time-varying neural systems.

Ax Zelin Wan, Jin-Hee Cho, Mu Zhu, Ahmed H. Anwar, Charles Kamhoua, Munindar P. Singh 3/24/2026

Cyber Deception for Mission Surveillance via Hypergame-Theoretic Deep Reinforcement Learning

Deep reinforcement learning approach using cyber deception via honey drones to defend UAVs against DoS attacks.

Ax Aarush Sinha, Rahul Seetharaman, Aman Bansal 3/24/2026

ECI: Effective Contrastive Information to Evaluate Hard-Negatives

ECI method for identifying effective hard negatives to improve dense retrieval model training efficiency.

Ax Abhinaba Basu 3/24/2026

Structural Sensitivity in Compressed Transformers: Error Propagation, Lyapunov Stability, and Formally Verified Bounds

Analysis of sensitivity and error propagation in compressed transformers across 5 architectures, mapping compression vulnerability hierarchy.

Ax Daniel Autenrieth 3/24/2026

How AI Systems Think About Education: Analyzing Latent Preference Patterns in Large Language Models

Analysis of educational alignment and preference patterns in GPT-5.1 across eight educational-theoretical dimensions.

Ax Rushikesh Deotale, Adithya Srinivasan, Yuan Tian, Tianyi Zhang, Pavlos Vlachos, Hector Gomez 3/24/2026

ALL-FEM: Agentic Large Language models Fine-tuned for Finite Element Methods

ALL-FEM: LLM agents fine-tuned for finite element method analysis and code generation in computational engineering.

Ax Jinquan Zheng, Jia Yuan, Jiacheng Yao, Chenyang Gu, Pujun Zheng, Guoxiu He 3/24/2026

Mitigating Selection Bias in Large Language Models via Permutation-Aware GRPO

Permutation-Aware GRPO training method to mitigate selection bias in LLMs on multiple-choice and pairwise evaluation tasks.

Ax Yunhai Hu, Junwei Zhou, Yumo Cao, Yitao Long, Yiwei Xu, Qiyi Jiang, Weiyao Wang, Xiaoyu Cao, Zhen Sun, Yiran Zou, Nan Du 3/24/2026

DSL-R1: From SQL to DSL for Training Retrieval Agents across Structured and Unstructured Data with Reinforcement Learning

DSL-R1 framework training retrieval agents via reinforcement learning across structured and unstructured data using domain-specific language.

Ax Tin Van Huynh, Kiet Van Nguyen, Ngan Luu-Thuy Nguyen 3/24/2026

ViCLSR: A Supervised Contrastive Learning Framework with Natural Language Inference for Natural Language Understanding Tasks

Supervised contrastive learning framework with natural language inference for Vietnamese language understanding in low-resource settings.

Ax Tasmay Pankaj Tibrewal, Pritish Saha, Ankit Meda, Kunal Singh, Pradeep Moturi 3/24/2026

Mixture of Chapters: Scaling Learnt Memory in Transformers

Mixture of Experts-inspired sparse memory banks for transformers, enabling scalable knowledge storage and retrieval via cross-attention.

Ax Long Xu, Junping Guo, Jianbo Zhao, Jianbo Lu, Yuzhong Peng 3/24/2026

DMMRL: Disentangled Multi-Modal Representation Learning via Variational Autoencoders for Molecular Property Prediction

Variational autoencoder approach for disentangled multi-modal representation learning in molecular property prediction for drug discovery.

Ax Peihan Ye, Alfreds Lapkovskis, Alaa Saleh, Qiyang Zhang, Praveen Kumar Donta 3/24/2026

NeSy-Edge: Neuro-Symbolic Trustworthy Self-Healing in the Computing Continuum

Neuro-symbolic framework for trustworthy self-healing in distributed edge computing environments with heterogeneous resources.

Ax Octavian Untila 3/24/2026

Emergent Formal Verification: How an Autonomous AI Ecosystem Independently Discovered SMT-Based Safety Across Six Domains

Autonomous AI ecosystem independently discovered SMT-based formal verification across six AI safety domains including LLM code, agent APIs, and smart contracts.

Ax Feng Liu, Jian Xu, Xin Cui, Xinghao Wang, Zijie Guo, Jiong Wang, S. Mostafa Mousavi, Xinyu Gu, Hao Chen, Ben Fei, Lihua Fang, Fenghua Ling, Zefeng Li, Lei Bai 3/24/2026

TRACE: A Multi-Agent System for Autonomous Physical Reasoning in Seismological Science

Multi-agent AI system for autonomous physical reasoning in seismology, integrating catalogs and models for earthquake mechanism inference.

Ax Zhiqiang He 3/24/2026

Rethinking Plasticity in Deep Reinforcement Learning

Investigation of plasticity loss mechanisms in deep reinforcement learning, proposing optimization-centric hypothesis to explain neural network adaptation failures.

Ax Kwanyoung Kim, Byeongsu Sim 3/24/2026

Reward Sharpness-Aware Fine-Tuning for Diffusion Models

Reward sharpness-aware fine-tuning approach to prevent reward hacking in diffusion model alignment via reinforcement learning.

Ax Andrei Baroian, Rutger Berger 3/24/2026

Prompt replay: speeding up grpo with on-policy reuse of high-signal prompts

Prompt Replay method for efficient GRPO training that reuses high-signal prompts to reduce compute waste in LLM reinforcement learning.

Ax Miryala Sathvika, Rudra Dhar, Karthik Vaidhyanathan 3/24/2026

LLM-based Automated Architecture View Generation: Where Are We Now?

Empirical evaluation of LLMs and agentic approaches for automated software architecture view generation from source code across 340 repositories.

Ax Sai Koneru, Jian Wu, Sarah Rajtmajer 3/24/2026

Context Selection for Hypothesis and Statistical Evidence Extraction from Full-Text Scientific Articles

NLP method for extracting hypotheses and statistical evidence from scientific papers using sequential full-text processing.

Ax Qiuchi Xiang, Haoxuan Qu, Hossein Rahmani, Jun Liu 3/24/2026

Is Monitoring Enough? Strategic Agent Selection For Stealthy Attack in Multi-Agent Discussions

Study of adversarial attacks on multi-agent LLM discussions under monitoring, exploring vulnerabilities in collaborative agent systems.

Ax Di Lu, Yongzhi Liao, Xutong Mu, Lele Zheng, Ke Cheng, Xuewen Dong, Yulong Shen, Jianfeng Ma 3/24/2026

When Convenience Becomes Risk: A Semantic View of Under-Specification in Host-Acting Agents

Research on security risks in host-acting AI agents, focusing on semantic under-specification in goal specifications and insufficient safety boundaries.

Ax Zhongyang Li, Yaqian Li, Faming Fang, Rinyoichi Takezoe, Zi-Hao Bo, Cheng Qian, Mo Guang, Guixu Zhang, Kaiwen Long 3/24/2026

QMoP: Query Guided Mixture-of-Projector for Efficient Visual Token Compression

Query Guided Mixture-of-Projector method for efficient visual token compression in multimodal LLMs, adaptively handling visual-text alignment across diverse scenarios.

Ax Zihan Fang, Qianru Wang, Haonan An, Zheng Lin, Yiqin Deng, Xianhao Chen, Yuguang Fang 3/24/2026

Aggregation Alignment for Federated Learning with Mixture-of-Experts under Data Heterogeneity

Federated learning framework for fine-tuning Mixture-of-Experts based LLMs on distributed privacy-sensitive data with aggregation alignment across heterogeneous clients.

Ax Pranav Hemanth, Sampriti Saha 3/24/2026

Conversation Tree Architecture: A Structured Framework for Context-Aware Multi-Branch LLM Conversations

Conversation Tree Architecture framework addressing logical context poisoning in multi-topic LLM conversations through structured branching conversation interfaces.

Ax Zongjie Li, Chaozheng Wang, Yuchong Xie, Pingchuan Ma, Shuai Wang 3/24/2026

WARBENCH: A Comprehensive Benchmark for Evaluating LLMs in Military Decision-Making

WARBENCH comprehensive benchmark evaluating LLM performance in military decision-making with International Humanitarian Law constraints and edge computing limitations.

Ax Soudeep Ghoshal, Sandipan Chakraborty, Pradipto Chowdhury, Himanshu Buckchash 3/24/2026

Fusing Memory and Attention: A study on LSTM, Transformer and Hybrid Architectures for Symbolic Music Generation

Comparative study of LSTM, Transformer, and hybrid architectures for symbolic music generation analyzing local and global structural coherence.

Ax Minjong Cheon 3/24/2026

Sonny: Breaking the Compute Wall in Medium-Range Weather Forecasting

Sonny, a compute-efficient deep learning model for medium-range weather forecasting competing with operational numerical systems.

Ax Zhengxian Wu, Kai Shi, Chuanrui Zhang, Zirui Liao, Jun Yang, Ni Yang, Qiuying Peng, Luyuan Zhang, Hangrui Xu, Tianhuang Su, Zhenyu Yang, Haonan Lu, Haoqian Wang 3/24/2026

When Models Judge Themselves: Unsupervised Self-Evolution for Multimodal Reasoning

Unsupervised self-evolution training framework for multimodal reasoning in LLMs achieving performance improvements without human annotations or teacher models.

Ax Trung V. Phan, Thomas Bauschert 3/24/2026

DeepXplain: XAI-Guided Autonomous Defense Against Multi-Stage APT Campaigns

DeepXplain framework integrating explainable AI with deep reinforcement learning for autonomous defense against multi-stage APT cyber attacks.

Ax Vinay Sharma, Manish Jain 3/24/2026

enhancing reasoning accuracy in large language models during inference time

Study of inference-time techniques to improve LLM reasoning accuracy through self-consistency, stochastic decoding, and other strategies without additional training.

Ax Xiaozhe Li, Tianyi Lyu, Siyi Yang, Yizhao Yang, Yuxi Gong, Jinxuan Huang, Ligao Zhang, Zhuoyi Huang, Qingwen Liu 3/24/2026

COINBench: Moving Beyond Individual Perspectives to Collective Intent Understanding

COINBench benchmark for evaluating LLM ability to understand collective intent by extracting consensus and resolving contradictions from multi-source discussions.

Ax Oussama Zekri, Th\'eo Uscidda, Nicolas Boull\'e, Anna Korba 3/24/2026

Generalized Discrete Diffusion from Snapshots

Generalized Discrete Diffusion from Snapshots framework enabling flexible discrete diffusion modeling over large state spaces with arbitrary corruption dynamics.

Ax K. M. Jubair Sami, Dipto Sumit, Ariyan Hossain, Farig Sadeque 3/24/2026

Benchmarking Bengali Dialectal Bias: A Multi-Stage Framework Integrating RAG-Based Translation and Human-Augmented RLAIF

Two-phase framework using RAG-based translation and RLAIF to evaluate and measure dialectal bias in LLM question-answering across Bengali dialects.

Ax Mariela M. Nina, Caio Veloso Costa, Lilian Berton, Didier A. Vega-Oliveros 3/24/2026

Efficient Fine-Tuning Methods for Portuguese Question Answering: A Comparative Study of PEFT on BERTimbau and Exploratory Evaluation of Generative LLMs

Systematic evaluation of Parameter-Efficient Fine-Tuning and quantization techniques for Portuguese question answering on BERTimbau, addressing low-resource language accessibility.

Ax Shuai Wang, Yinan Yu, Earl Barr, Dhasarathy Parthasarathy 3/24/2026

LLM-Powered Workflow Optimization for Multidisciplinary Software Development: An Automotive Industry Case Study

Research on using LLMs to optimize multidisciplinary software development workflows in automotive industry, automating coordination between domain experts and developers.

Ax Shuai Wang, Yinan Yu 3/24/2026

KG-Hopper: Empowering Compact Open LLMs with Knowledge Graph Reasoning via Reinforcement Learning

KG-Hopper enhances compact open-source LLMs with knowledge graph reasoning via reinforcement learning for improved knowledge-intensive question answering.

Ax Yubo Li, Ramayya Krishnan, Rema Padman 3/24/2026

When Documents Disagree: Measuring Institutional Variation in Transplant Guidance with Retrieval-Augmented Language Models

Framework using retrieval-augmented language models to measure institutional variation in patient education materials across 102 organ transplant centers.

Ax James Wedgwood, Aashiq Muhamed, Mona T. Diab, Virginia Smith 3/24/2026