Ax Akshay K. Jagadish, Milena Rmus, Kristin Witte, Marvin Mathony, Marcel Binz, Eric Schulz 3/24/2026

Can we automatize scientific discovery in the cognitive sciences?

Proposes automating scientific discovery in cognitive sciences using agentic AI to explore computational models faster than manual researcher-driven cycles.

Ax Benedikt Hornig, Reuth Mirsky 3/24/2026

The Intelligent Disobedience Game: Formulating Disobedience in Stackelberg Games and Markov Decision Processes

Formalizes intelligent disobedience in shared autonomy using Stackelberg games and MDPs, modeling when automated assistants should override human instructions for safety.

Ax Erich Studerus, Vivienne Jia Zhong, Stephan Vonschallen 3/24/2026

A Framework for Low-Latency, LLM-driven Multimodal Interaction on the Pepper Robot

Framework for integrating LLMs into Pepper robot for low-latency multimodal interaction, addressing cascaded pipeline latency and improving agentic control.

Ax Ziquan Wang, Zhongqi Lu 3/24/2026

Knowledge Boundary Discovery for Large Language Models

Proposes Knowledge Boundary Discovery, an RL framework to map LLM knowledge boundaries by generating questions the model can and cannot confidently answer.

Ax Ye Tian, Jingyi Zhang, Zihao Wang, Xiaoyuan Ren, Xiaofan Yu, Onat Gungor, Tajana Rosing 3/24/2026

KLDrive: Fine-Grained 3D Scene Reasoning for Autonomous Driving based on Knowledge Graph

Presents KLDrive, using knowledge graphs for fine-grained 3D scene reasoning in autonomous driving with LLMs, reducing hallucinations and improving reasoning transparency.

Ax Jianing Wang, Jianfei Zhang, Qi Guo, Linsen Guo, Rumei Li, Chao Zhang, Chong Peng, Cunguang Wang, Dengchang Zhao, Jiarong Shi, Jingang Wang, Liulin Feng, Mengxia Shen, Qi Li, Shengnan An, Shun Wang, Wei Shi, Xiangyu Xi, Xiaoyu Li, Xuezhi Cao, Yi Lu, Yunke Zhao, Zhengyu Chen, Zhimin Lin, Wei Wang, Peng Pei, Xunliang Cai 3/24/2026

LongCat-Flash-Prover: Advancing Native Formal Reasoning via Agentic Tool-Integrated Reinforcement Learning

Open-source 560B MoE model advancing formal reasoning in Lean4 via agentic tool-integrated reinforcement learning, decomposing tasks into auto-formalization, sketching, and proving.

Ax Zhuojie Yang, Wentao Wan, Keze Wang 3/24/2026

ORACLE: Optimizing Reasoning Abilities of Large Language Models via Constraint-Led Synthetic Data Elicitation

Proposes ORACLE, a method for generating high-quality synthetic reasoning data to train LLMs by validating intermediate reasoning steps beyond final answer correctness.

Ax Zihui Chen, Yuling Wang, Pengfei Jiao, Kai Wu, Xiao Wang, Xiang Ao, Dalin Zhang 3/24/2026

Can LLMs Fool Graph Learning? Exploring Universal Adversarial Attacks on Text-Attributed Graphs

Studies adversarial attacks on text-attributed graphs combining GNNs and language models, exploring vulnerabilities in graph learning systems.

Ax Leonid Ugadiarov, Yuri Kuratov, Aleksandr Panov, Alexey Skrynnik 3/24/2026

Revisiting Tree Search for LLMs: Gumbel and Sequential Halving for Budget-Scalable Reasoning

Addresses scaling failures in AlphaZero-style tree search for LLM reasoning, proposes ReSC using Gumbel sampling and sequential halving for budget-scalable inference.

Ax Haoyu Qiao, Hao Zhang, Shanwen Mao, Siyao Cheng, Jie Liu 3/24/2026

ConsRoute:Consistency-Aware Adaptive Query Routing for Cloud-Edge-Device Large Language Models

Introduces ConsRoute, a semantic-aware adaptive routing framework for distributing LLM queries across cloud-edge-device tiers to reduce latency and inference costs.

Ax Yu Luo, Rongchen Gao, Lu Teng, Xidao Wen, Jiamin Jiang, Qingliang Zhang, Yongqian Sun, Shenglin Zhang, Jiasong Feng, Tong Liu, Wenjie Zhang, Dan Pei 3/24/2026

Graph of States: Solving Abductive Tasks with Large Language Models

Proposes Graph of States framework for LLMs to perform abductive reasoning through structured state representation and explicit state control, addressing gaps in logical reasoning.

Ax Zachary F. Mainen 3/24/2026

The Library Theorem: How External Organization Governs Agentic Reasoning Capacity

Formalizes transformer context windows as I/O pages, proving tool-augmented agents with indexed external memory achieve exponential retrieval cost improvements over sequential scanning.

Ax Pantea Karimi, Kimia Noorbakhsh, Mohammad Alizadeh, Hari Balakrishnan 3/24/2026

Improving Coherence and Persistence in Agentic AI for System Optimization

Approach improving LLM agent coherence for system optimization by addressing evolutionary bias and coherence ceiling.

Ax Seth Dobrin, Lukasz Chmiel 3/24/2026

ARYA: A Physics-Constrained Composable & Deterministic World Model Architecture

Physics-constrained composable world model architecture ARYA designed for deterministic planning and AI safety.

Ax Dongyoung Kim, Sumin Park, Woomin Song, Seungku Kim, Taeyoung Kim, Huiwon Jang, Jinwoo Shin, Jaehyung Kim, Younggyo Seo 3/24/2026

RoboAlign: Learning Test-Time Reasoning for Language-Action Alignment in Vision-Language-Action Models

RoboAlign approach improving embodied reasoning in vision-language-action models through test-time reasoning optimization.

Ax Ulisses Braga-Neto 3/24/2026

The AI Scientific Community: Agentic Virtual Lab Swarms

Framework for AI scientific research using swarms of virtual lab agents enabling decentralized collaborative exploration.

Ax Liang Ding 3/24/2026

AgentHER: Hindsight Experience Replay for LLM Agent Trajectory Relabeling

AgentHER adapts hindsight experience replay to recover training signal from failed LLM agent trajectories.

Ax Liang Ding 3/24/2026

AdaRubric: Task-Adaptive Rubrics for LLM Agent Evaluation

AdaRubric generates task-specific evaluation rubrics on-the-fly for LLM agent assessment with step-by-step feedback.

Ax Junkeun Yi, Damon Mosk-Aoyama, Baihe Huang, Ritu Gala, Charles Wang, Sugam Dipak Devare, Khushi Bhardwaj, Abhibha Gupta, Oleksii Kuchaiev, Jiantao Jiao, Jian Zhang, Venkat Srinivasan 3/24/2026

PivotRL: High Accuracy Agentic Post-Training at Low Compute Cost

PivotRL framework improving agentic post-training efficiency by combining supervised fine-tuning with reinforcement learning.

Ax Johnathan Sun, Andrew Zhang 3/24/2026

Persona Vectors in Games: Measuring and Steering Strategies via Activation Vectors

Method for measuring and steering LLM strategic behavior in games using activation vectors for personality traits.

Ax Anthony T. Nixon 3/24/2026

The Myhill-Nerode Theorem for Bounded Interaction: Canonical Abstractions via Agent-Bounded Indistinguishability

Theoretical framework extending Myhill-Nerode theorem to bounded agents in finite POMDPs for canonical abstractions.

Ax Gregory M. Ruddell 3/24/2026

Silent Commitment Failure in Instruction-Tuned Language Models: Evidence of Governability Divergence Across Architectures

Study of error detectability in instruction-tuned LLM agents, proposing governability metric for agent security.

Ax Shuai Wang, Dhasarathy Parthasarathy, Robert Feldt, Yinan Yu 3/24/2026

DomAgent: Leveraging Knowledge Graphs and Case-Based Reasoning for Domain-Specific Code Generation

Agent architecture combining knowledge graphs and case-based reasoning for domain-specific code generation with LLMs.

Ax Taejin Park 3/24/2026

Behavioural feasible set: Value alignment constraints on AI decision support

Formalizes vendor value alignment constraints in AI decision support systems as behavioral feasible sets.

Ax Cosimo Spera 3/24/2026

Safety as Computation: Certified Answer Reuse via Capability Closure in Task-Oriented Dialogue

Task-oriented dialogue system approach treating safety certification as computational primitive for answer reuse.

Ax Hehai Lin, Yu Yan, Zixuan Wang, Bo Xu, Sudong Wang, Weiquan Huang, Ruochen Zhao, Minzhi Li, Chengwei Qin 3/24/2026

Unified-MAS: Universally Generating Domain-Specific Nodes for Empowering Automatic Multi-Agent Systems

Framework for automatically generating domain-specific agent nodes to improve multi-agent systems in specialized fields.

Ax Xinyu Zhang 3/24/2026

Stabilizing Iterative Self-Training with Verified Reasoning via Symbolic Recursive Self-Alignment

Neuro-symbolic approach stabilizing iterative self-training in LLMs by preventing recursive drift through verified reasoning.

Ax Zhongyi Li, Wan Tian, Yikun Ban, Jinju Chen, Huiming Zhang, Yang Liu, Fuzhen Zhuang 3/24/2026

Counterfactual Credit Policy Optimization for Multi-Agent Collaboration

Framework for solving credit assignment in collaborative multi-agent LLMs via counterfactual policy optimization to reduce free-riding.

Ax Zhongyi Li, Wan Tian, Jingyu Chen, Kangyao Huang, Huiming Zhang, Hui Yang, Tao Ren, Jinyang Jiang, Yijie Peng, Yikun Ban, Fuzhen Zhuang 3/24/2026

Adaptive Robust Estimator for Multi-Agent Reinforcement Learning

Research on robust credit assignment for multi-agent LLM collaboration using reinforcement learning with heavy-tailed reward handling.

Ax Qihui Zhu, Shouwei Ruan, Xiao Yang, Hao Jiang, Yao Huang, Shiji Zhao, Hanwei Fan, Hang Su, Xingxing Wei 3/24/2026

Mind over Space: Can Multimodal Large Language Models Mentally Navigate?

Study of multimodal LLM spatial reasoning capabilities, showing limitations in mental navigation and long-range spatiotemporal planning despite embodied agent deployment.

Ax Sheng Liu, Long Chen, Zeyun Zhao, Qinglin Gou, Qingyue Wei, Arjun Masurkar, Kevin M. Spiegler, Philip Kuball, Stefania C. Bray, Megan Bernath, Deanna R. Willis, Jiang Bian, Lei Xing, Eric Topol, Kyunghyun Cho, Yu Huang, Ruogu Fang, Narges Razavian, James Zou 3/24/2026

A Multidisciplinary AI Board for Multimodal Dementia Characterization and Risk Assessment

Multi-agent AI team (Cerebra) coordinating specialized agents for clinical decision support on heterogeneous patient data across EHR, notes, and medical imaging.

Ax Alexandra Bazarova, Andrei Volodichev, Daria Kotova, Alexey Zaytsev 3/24/2026

INTRYGUE: Induction-Aware Entropy Gating for Reliable RAG Uncertainty Estimation

Method for improving uncertainty quantification in RAG systems by addressing entropy-based estimation failures caused by internal tug-of-war between induction and grounding.

Ax Ankush Agarwal, Harsh Vishwakarma, Suraj Nagaje, Chaitanya Devaguptapu 3/24/2026

EnterpriseLab: A Full-Stack Platform for developing and deploying agents in Enterprises

Full-stack platform for enterprise AI agent development unifying tool integration, data generation, and training in closed-loop framework with privacy preservation.

Ax Yiliang Song, Hongjun An, Jiangan Chen, Xuanchen Yan, Huan Song, Jiawei Shao, Xuelong Li 3/24/2026

Silicon Bureaucracy and AI Test-Oriented Education: Contamination Sensitivity and Score Confidence in LLM Benchmarks

Critical analysis of LLM benchmark contamination and overfitting, arguing benchmark scores conflate test-oriented competence with genuine capability generalization.

Ax Mohammad Asadi, Jack W. O'Sullivan, Fang Cao, Tahoura Nedaee, Kamyar Fardi, Fei-Fei Li, Ehsan Adeli, Euan Ashley 3/24/2026

Mirage The Illusion of Visual Understanding

Research revealing gaps in multimodal AI systems' visual understanding mechanisms, showing frontier models generate plausible but unsupported reasoning traces.

Ax Yicai Xing 3/24/2026

AI Token Futures Market: Commoditization of Compute and Derivatives Contract Design

Analysis of AI tokens as commodities and design of derivative contracts for inference compute, treating token consumption as infrastructure raw materials.

Ax Neelmani Vispute 3/24/2026

Reasoning Provenance for Autonomous AI Agents: Structured Behavioral Analytics Beyond State Checkpoints and Execution Traces

Framework for analyzing reasoning behavior and decision-making patterns in autonomous AI agents through structured behavioral analytics and operational tooling.

Ax Mohammad Asadi, Tahoura Nedaee, Jack W. O'Sullivan, Euan Ashley, Ehsan Adeli 3/24/2026

Deterministic Hallucination Detection in Medical VQA via Confidence-Evidence Bayesian Gain

Bayesian method for detecting hallucinations in medical VQA using confidence-evidence gains without multiple stochastic generations.

Ax Hunmin Do, Taejun Yoon, Kiyong Jung 3/24/2026

MIND: Multi-agent inference for negotiation dialogue in travel planning

MIND: Multi-agent framework using Theory of Mind for realistic consensus-building in travel planning with heterogeneous preferences.

Ax Jinhui Ren, Huaiming Li, Yabin Liu, Tao Li, Zhaokun Liu, Yujia Liang, Zengle Ge, Chufan Wu, Xiaomin Yuan, Danyu Liu, Annan Li, Jianmin Wu 3/24/2026

A Blueprint for Self-Evolving Coding Agents in Vehicle Aerodynamic Drag Prediction

Self-evolving coding agent blueprint for discovering surrogate ML pipelines for vehicle aerodynamic drag prediction.

Ax Xi Wang, Xu Yang, Donghao Sun, Cheng Deng 3/24/2026

Compensating Visual Insufficiency with Stratified Language Guidance for Long-Tail Class Incremental Learning

Method leveraging LLM language guidance to address tail class scarcity in long-tail class incremental learning scenarios.

Ax Shuo Wang, Ziyu Chen, Ming Tang 3/24/2026

CurvZO: Adaptive Curvature-Guided Sparse Zeroth-Order Optimization for Efficient LLM Fine-Tuning

CurvZO: Efficient LLM fine-tuning via adaptive curvature-guided zeroth-order optimization with reduced memory overhead.

Ax Andreas Sauter, Yuyue Zhao, Jacopo Urbani, Wenxiang Hu, Zaiqiao Meng, Lun Zhou, Xiaohui Yan, Yougang Lyu 3/24/2026

EvoIdeator: Evolving Scientific Ideas through Checklist-Grounded Reinforcement Learning

EvoIdeator: RL approach for autonomous scientific idea generation using checklist-grounded refinement instead of scalar rewards.

Ax Yiling Wu 3/24/2026

The Reasoning Error About Reasoning: Why Different Types of Reasoning Require Different Representational Structures

Framework identifying four structural properties of representational systems demanded by different types of reasoning.

Ax Yiling Wu 3/24/2026

The Presupposition Problem in Representation Genesis

Philosophical analysis of how LLMs differ from prior cognitive systems in representation genesis without clear transitions.

Ax Susana Nunes, Tiago Guerreiro, Catia Pesquita 3/24/2026

Agentic Personas for Adaptive Scientific Explanations with Knowledge Graphs

Framework for agentic personas that provide adaptive scientific explanations using knowledge graphs based on expert goals.

Ax Aryan Kasat, Smriti Singh, Aman Chadha, Vinija Jain 3/24/2026

Reasoning or Rhetoric? An Empirical Analysis of Moral Reasoning Explanations in Large Language Models

Empirical analysis testing whether LLMs exhibit genuine moral reasoning or produce superficially similar outputs through alignment training.

Ax Naoshi Uchihira 3/24/2026

Tacit Knowledge Management with Generative AI: Proposal of the GenAI SECI Model

GenAI SECI Model: Framework for using generative AI to manage tacit knowledge in organizations.

Ax Shuying Chen, Sen Cui, Zhong Cao 3/24/2026

Guideline-grounded retrieval-augmented generation for ophthalmic clinical decision support

Oph-Guid-RAG: Multimodal RAG system for ophthalmology clinical decision support retrieving guideline pages as evidence units.

Ax Caio Azevedo, Stefano Sabatini, Sascha Hornauer, Fabien Moutarde 3/24/2026

Future-Interactions-Aware Trajectory Prediction via Braid Theory

Braid theory approach to multi-agent trajectory prediction for autonomous vehicles without extensive computation or heuristics.