Ax Mohd Sameen Chishti, Damilare Peter Oyinloye, Jingyue Li 5/5/2026

AgentReputation: A Decentralized Agentic AI Reputation Framework

Decentralized reputation framework for agentic AI marketplaces handling strategic optimization and task context transfer.

Ax Shubham Kumar, Narendra Ahuja 5/5/2026

Minimal, Local, Causal Explanations for Jailbreak Success in Large Language Models

Methods for explaining jailbreak vulnerabilities in LLMs through local causal analysis of model representations.

Ax Kaituo Zhang, Zhen Xiong, Mingyu Zhong, Zhimeng Jiang, Zhouyuan Yuan, Zhecheng Li, Ying Lin 5/5/2026

Are Tools All We Need? Unveiling the Tool-Use Tax in LLM Agents

Analysis of tool-use overhead in LLM agents, showing semantic distractors can degrade performance vs. chain-of-thought reasoning.

Ax Abdulhady Abas Abdullah, Fatemeh Daneshfar, Seyedali Mirjalili, Mourad Oussalah 5/5/2026

TUR-DPO: Topology- and Uncertainty-Aware Direct Preference Optimization

TUR-DPO method for LLM alignment that improves Direct Preference Optimization by accounting for preference topology and uncertainty.

Ax Sydney Johns, Heng Jin, Chaoyu Zhang, Y. Thomas Hou, Wenjing Lou 5/5/2026

ARMOR 2025: A Military-Aligned Benchmark for Evaluating Large Language Model Safety Beyond Civilian Contexts

Safety benchmark for evaluating LLMs in military/defense contexts with doctrinal standards for decision support systems.

Ax Frederik Hytting J{\o}rgensen, Sebastian Weichwald, Lewis Hammond 5/5/2026

Causal Foundations of Collective Agency

Theoretical framework studying when multiple agents form a unified collective agent with emergent capabilities distinct from individuals.

Ax Tiejin Chen, Ahmadreza Moradipari, Kyungtae Han, Hua Wei, Nejib Ammar 5/5/2026

Agentic AI for Trip Planning Optimization Application

System for optimizing trip planning for intelligent vehicles considering travel time, energy consumption, and traffic using agentic AI.

Ax Yuxuan Gao, Megan Wang, Yi Ling Yu 5/5/2026

Token Arena: A Continuous Benchmark Unifying Energy and Cognition in AI Inference

TokenArena continuous benchmark measures AI inference endpoints across speed, latency, and cost metrics at granular deployment levels.

Ax Ranit Karmakar, Jayita Chatterjee 5/5/2026

AgentFloor: How Far Up the tool use Ladder Can Small Open-Weight Models Go?

AgentFloor benchmark evaluates which agent workflow tasks require large models vs. smaller models, introducing 30-task capability ladder for routing decisions.

Ax Sen Cui, Jingheng Ma 5/5/2026

Physically Native World Models: A Hamiltonian Perspective on Generative World Modeling

Study of world models for embodied AI and robotics using Hamiltonian mechanics, unifying 2D video, 3D scene, and latent prediction approaches.

Ax Haotian Zhao, Yuxin Zhang, Songlin Zhou, Stephen S. -T. Yau, Wenyu Zhang, Lun Tian, Tianshu Zhu, Yifeng Huang, Yucheng Zeng, Jingnan Gu, Daxiang Dong, Jianmin Wu 5/5/2026

AEM: Adaptive Entropy Modulation for Multi-Turn Agentic Reinforcement Learning

Research on Adaptive Entropy Modulation for training LLM agents via reinforcement learning, addressing credit assignment in multi-turn tasks through intermediate supervision.

Ax Jinkun Liu, Haohan Chi, Lingfeng Zhang, Yifan Xie, YuAn Wang, Long Chen, Hangjun Ye, Xiaoshuai Hao, Wenbo Ding 5/5/2026

Thinking in Text and Images: Interleaved Vision--Language Reasoning Traces for Long-Horizon Robot Manipulation

Interleaved vision-language reasoning traces for robot manipulation, combining text-based planning and visual grounding for long-horizon task execution.

Ax Yan Zhang, Daiqing Wu, Huawen Shen, Yu Zhou, Can Ma 5/5/2026

Learn where to Click from Yourself: On-Policy Self-Distillation for GUI Grounding

On-policy self-distillation method for GUI grounding agents using reinforcement learning to improve autonomous UI interaction without expensive rollouts.

Ax Qinyuan Wu, Soumi Das, Mahsa Amani, Arijit Nag, Seungeon Lee, Krishna P. Gummadi, Abhilasha Ravichander, Muhammad Bilal Zafar 5/5/2026

To Call or Not to Call: A Framework to Assess and Optimize LLM Tool Calling

Framework for evaluating when LLM agents should call external tools vs rely on internal knowledge, analyzing tool-use decision quality and optimization.

Ax Theodore Papamarkou, Pierre Alquier, Matthias Bauer, Wray Buntine, Andrew Davison, Gintare Karolina Dziugaite, Maurizio Filippone, Andrew Y. K. Foong, Vincent Fortuin, Dimitris Fouskakis, Jes Frellsen, Eyke H\"ullermeier, Theofanis Karaletsos, Mohammad Emtiyaz Khan, Nikita Kotelevskii, Salem Lahlou, Yingzhen Li, Fang Liu, Clare Lyle, Thomas M\"ollenhoff, Konstantina Palla, Maxim Panov, Yusuf Sale, Kajetan Schweighofer, Artem Shelmanov, Siddharth Swaroop, Martin Trapp, Willem Waegeman, Andrew Gordon Wilson, Alexey Zaytsev 5/5/2026

Position: agentic AI orchestration should be Bayes-consistent

Position paper arguing agentic AI systems should use Bayesian approaches for the control layer orchestrating LLMs and tools under uncertainty.

Ax Garvin Kruthof 5/5/2026

Models Recall What They Violate: Constraint Adherence in Multi-Turn LLM Ideation

DriftBench benchmark evaluating constraint adherence across 2,146 multi-turn LLM interactions from 7 models, testing fidelity to original objectives in scientific ideation tasks.

Ax Pragya Sharma, Hang Qiu, Mani Srivastava 5/5/2026

Cloud Is Closer Than It Appears: Revisiting the Tradeoffs of Distributed Real-Time Inference

Research on distributed vs on-device inference tradeoffs for real-time deep learning in cyber-physical systems, analyzing computational and latency constraints.

Ax Michael Chertkov 5/5/2026

Mean-Field Path-Integral Diffusion: From Samples to Interacting Agents

Mean-Field Path-Integral Diffusion: generative model where samples act as interacting agents coordinating via population statistics for efficient sampling.

Ax Md Sirajul Islam, Isabelle G Chapman, N I Md Ashafuddula, Xu Yuan, Li Chen, Nian-Feng Tzeng, Klara Nahrstedt 5/5/2026

FedACT: Concurrent Federated Intelligence across Heterogeneous Data Sources

FedACT: Federated learning framework enabling concurrent training of multiple ML tasks across heterogeneous decentralized devices.

Ax Roman Smirnov 5/5/2026

Exploring LLM biases to manipulate AI search overview

Analysis of LLM biases in search overview systems and methods to manipulate AI-generated search result summaries.

Ax Siyang Li, Yize Chen, Zijie Zhu, Yuxin Pan, Yan Guo, Ming Huang, Hui Xiong 5/5/2026

TimeRFT: Stimulating Generalizable Time Series Forecasting for TSFMs via Reinforcement Finetuning

TimeRFT: Reinforcement finetuning approach for time series foundation models to improve adaptation to downstream forecasting tasks.

Ax Woody Haosheng Gan, William Held, Diyi Yang 5/5/2026

Putting HUMANS first: Efficient LAM Evaluation with Human Preference Alignment

Efficient evaluation methodology for large audio models using minimal subsets aligned with human preferences across 40 tasks.

Ax Yu Shen, Shiyang Liu, Qihang He, Yihang Cheng, Haining Xie, Zhiming He, Huahua Fan, Xianzhi Tan, Teng Ma, Shaoquan Zhang, Danqing Huang, Fan Jiang, Yang Li, Chongqing Zhao, Peng Chen, Jie Jiang, Bin Cui 5/5/2026

SiriusHelper: An LLM Agent-Based Operations Assistant for Big Data Platforms

SiriusHelper: LLM agent-based operations assistant for big data platforms with RAG and knowledge retrieval optimization.

Ax Diego F. Cuadros, Abdoul-Aziz Maiga 5/5/2026

Ambient Persuasion in a Deployed AI Agent: Unauthorized Escalation Following Routine Non-Adversarial Content Exposure

Safety incident report: deployed AI agent escalated privileges and installed unauthorized components after exposure to routine content.

Ax Wentao Chen, Jingtang Chen, Mingjian Fu, Tiantian Li, Youfeng Su, Wenxi Liu, Yuanlong Yu 5/5/2026

Dynamic-TD3: A Novel Algorithm for UAV Path Planning with Dynamic Obstacle Trajectory Prediction

Deep reinforcement learning algorithm for UAV path planning with dynamic obstacle prediction and safety constraints.

Ax Yiyang Wei, Tingyu Song, Siyue Zhang, Yilun Zhao 5/5/2026

A Survey of Reasoning-Intensive Retrieval: Progress and Challenges

Survey of reasoning-intensive retrieval systems integrating LLM reasoning capabilities across IR pipeline from benchmarks to rerankers.

Ax Ricardo Luna Gutierrez, Sahand Ghorbanpour, Ejaz Rahman, Varchas Gopalaswamy, Riccardo Betti, Vineet Gundecha, Aarne Lees, Soumyendu Sarkar 5/5/2026

Human-in-the-Loop Meta Bayesian Optimization for Fusion Energy and Scientific Applications

Framework combining Bayesian optimization with human expertise for accelerating discovery in data-scarce scientific domains like fusion energy.

Ax Kenneth See, Xue Wen Tan 5/5/2026

Compliance-Aware Agentic Payments on Stablecoin Rails

Architecture for AI agents managing stablecoin payments with embedded compliance guardrails and signature-based authorization in regulated settings.

Ax Jiutian Zeng, Junjie Li, Chengwei Dai, Jie Liang, Zhaoyu Hu, Yiliang Zhang, Ziang Weng, Longtao Huang, Dongjie Zhang, Libin Dong, Yang Ge, Yuanda Wang, Kaiwen Lv Kacuila, Bingyu Zhu, Jing Wang, Jin Xu 5/5/2026

XekRung Technical Report

XekRung: Cybersecurity-focused large language model with specialized data synthesis pipelines and comprehensive training infrastructure.

Ax Shalini Sarode, Brian Moser, Joachim Folz, Federico Raue, Tobias Nauen, Stanislav Frolov, Andreas Dengel 5/5/2026

Hyperspherical Forward-Forward with Prototypical Representations

Novel reformulation of Forward-Forward algorithm using hyperspherical representations to improve inference efficiency for classification.

Ax Enzo S. N. Silva, Pablo B. Costa, Raphael C. Vlasman, Rosimeire P. Costa, Henrique L. P. Silva, Lucas F. A. O. Pellicer, Guilherme Rinaldo, Renato A. Almeida, Darian S. R. Rabbani, Cinthya O. Oestreich, Vinicius F. Carid\'a 5/5/2026

NorBERTo: A ModernBERT Model Trained for Portuguese with 331 Billion Tokens Corpus

NorBERTo: Portuguese language model based on ModernBERT architecture with 331B token training corpus and long-context support.

Ax Sichang Steven He, Calvin Ardi, Ramesh Govindan, Harsha V. Madhyastha 5/5/2026

DeGenTWeb: A First Look at LLM-dominant Websites

Empirical study examining prevalence of LLM-generated content on websites and evaluating detection methods with transparent methodology.

Ax Ishan Gupta, Pavlo Buryi 5/5/2026

How Frontier LLMs Adapt to Neurodivergence Context: A Measurement Framework for Surface vs. Structural Change in System-Prompted Responses

NDBench: benchmark measuring how frontier LLMs adjust outputs based on neurodivergence context in system prompts.

Ax Nhung Thi-Hong Duong, Mai Ngoc Ho, Tin Van Huynh, Kiet Van Nguyen 5/5/2026

ViLegalNLI: Natural Language Inference for Vietnamese Legal Texts

ViLegalNLI: first large-scale Vietnamese NLI dataset for legal domain with 42,012 premise-hypothesis pairs from statutory documents.

Ax Muhammad Dehan Al Kautsar, Saeed Almheiri, Momina Ahsan, Bilal Elbouardi, Younes Samih, Sarfraz Ahmad, Amr Keleg, Omar El Herraoui, Kareem Elzeky, Abed Alhakim Freihat, Mohamed Anwar, Zhuohan Xie, Junhong Liang, Mohammad Rustom Al Nasar, Preslav Nakov, Fajri Koto 5/5/2026

Cultural Benchmarking of LLMs in Standard and Dialectal Arabic Dialogues

Benchmark dataset (ArabCulture-Dialogue) evaluating LLM cultural reasoning in Standard and dialectal Arabic conversations across 13 countries.

Ax Debasis Dwibedy, Avyay Nishtala, Pranathi Mukku, D Snehaja 5/5/2026

Smart Profit-Aware Crop Advisory System: Kisan AI

Kisan AI: crop advisory system that optimizes for farmer profit rather than just biological yield by incorporating market prices.

Ax Mohammad Hossein Moslemi, Nima Hosseini Dashtbayaz, Zhimin Mei, Boyu Wang, Bissan Ghaddar 5/5/2026

Fair Dataset Distillation via Cross-Group Barycenter Alignment

Research on dataset distillation that preserves fairness across demographic groups through barycenter alignment techniques.

Ax Aur\'elien B\"uck-Kaeffer (McGill University, Mila - Quebec Artificial Intelligence Institute, Ubisoft La Forge), Sneheel Sarangi (McGill University, Mila - Quebec Artificial Intelligence Institute), Maximilian Puelma Touzel (McGill University, Universit\'e de Montr\'eal), Reihaneh Rabbany (McGill University, Mila - Quebec Artificial Intelligence Institute), Zachary Yang (McGill University, Mila - Quebec Artificial Intelligence Institute, Ubisoft La Forge), Jean-Fran\c{c}ois Godbout (Mila - Quebec Artificial Intelligence Institute, Universit\'e de Montr\'eal) 5/5/2026

The $\textit{Silicon Society}$ Cookbook: Design Space of LLM-based Social Simulations

Systematic analysis of design space for LLM-based social simulations, examining key design decisions and their consequences for simulating human behavior.

Ax Jugal Gajjar, Kamalasankari Subramaniakuppusamy 5/5/2026

RSAT: Structured Attribution Makes Small Language Models Faithful Table Reasoners

Method training small language models to perform table reasoning with cell-level citations using structured JSON output and faithfulness-based reward optimization.

Ax Jan Sobotka, Mustafa O. Karabag, Ufuk Topcu 5/5/2026

Why Do LLMs Struggle in Strategic Play? Broken Links Between Observations, Beliefs, and Actions

Analysis of LLM failures in strategic decision-making under incomplete information, identifying gaps between observations, beliefs, and actions in game-theoretic settings.

Ax Aviral Srivastava, Sourav Panda 5/5/2026

Attention Is Where You Attack

White-box adversarial attack on safety-aligned LLMs using attention redistribution to identify and exploit safety-critical attention heads with nonsemantic tokens.

Ax Junsun Choi, Sam Son, Sunjin Choi, Hansung Kim, Yakun Sophia Shao, Scott Shenker, Sylvia Ratnasamy, Borivoje Nikolic 5/5/2026

Rethinking Network Topologies for Cost-Effective Mixture-of-Experts LLM Serving

Systematic analysis of network infrastructure costs for serving Mixture-of-Experts LLMs, questioning necessity of expensive high-bandwidth networks for MoE deployment.

Ax Osmar Luiz Ferreira de Carvalho, Osmar Ab\'ilio de Carvalho J\'unior, Anesmar Olino de Albuquerque, Daniel Guerreiro e Silva 5/5/2026

Remote SAMsing: From Segment Anything to Segment Everything

Computer vision method extending Segment Anything Model 2 to remote sensing by addressing quality-coverage trade-offs and image tiling challenges for large-scale segmentation.

Ax Jatin Gupta, Akhil Sharma, Saransh Singhania, Ali Imam Abidi 5/5/2026

Retrieval-Augmented Reasoning for Chartered Accountancy

Research on using RAG with LLMs for Indian Chartered Accountancy tasks, addressing reliability and numerical reasoning challenges in jurisdiction-specific financial applications.

Ax Daniel Zhu, Zihan Wang, Jenny Bao, Jerry Wei 5/5/2026

Jailbroken Frontier Models Retain Their Capabilities

Study showing advanced jailbreaks on frontier models scale inversely with capability, with top jailbreaks imposing negligible performance tax.

Ax Sheza Munir, Ahanaf Rodoshi, Sumin Lee, Feiran Chang, Xujie Si, Syed Ishtiaque Ahmed 5/5/2026

Are You the A-hole? A Fair, Multi-Perspective Ethical Reasoning Framework

Neuro-symbolic framework using weighted MaxSAT for ethical reasoning aggregation across conflicting natural language judgments.

Ax Bingzheng Gan, Tianyi Zhang, Yusu Li, Jing Huang, Wei Shi, Yangkai Ding, Tao Yu 5/5/2026

Caracal: Causal Architecture via Spectral Mixing

Caracal replaces attention with O(L log L) Multi-Head Fourier module using FFT for efficient long-sequence LLM scaling.

Ax Hongbo Wen, Ying Li, Hanzhi Liu, Chaofan Shou, Yanju Chen, Yuan Tian, Yu Feng 5/5/2026

Semia: Auditing Agent Skills via Constraint-Guided Representation Synthesis

Semia audits LLM-driven agent skills via constraint-guided representation synthesis, testing hybrid artifact configurations for executable interfaces.

Ax Xingyu Hu, Kai Zhang, Jiancan Wu, Shuli Wang, Chi Wang, Wenshuai Chen, Yinhua Zhu, Haitao Wang, Xingxing Wang, Xiang Wang 5/5/2026

DynamicPO: Dynamic Preference Optimization for Recommendation

DynamicPO addresses preference optimization collapse in LLM-based recommendation systems by dynamically managing negative samples.

Ax Khizar Qureshi, Geoffrey Martin, Yifan Peng 5/5/2026

Budget-Aware Routing for Long Clinical Text

Budget-aware context selection for clinical text using knapsack-constrained subset selection to meet token cost and latency constraints.