Ax Ziang Ye, Wentao Shi, Yuxin Liu, Yu Wang, Zhengzhou Cai, Yaorui Shi, Qi Gu, Xunliang Cai, Fuli Feng 5/18/2026

Look Before You Leap: Autonomous Exploration for LLM Agents

Exploration Checkpoint Coverage metric for measuring autonomous exploration in LLM agents in unfamiliar environments.

Ax Parand A. Alamdari, Toryn Q. Klassen, Sheila A. McIlraith 5/18/2026

Formal Methods Meet LLMs: Auditing, Monitoring, and Intervention for Compliance of Advanced AI Systems

Formal methods combined with ML for auditing and monitoring AI systems throughout development lifecycle.

Ax Igor Bogdanov, Chung-Horng Lung, Thomas Kunz, Jie Gao, Adrian Taylor, Marzia Zaman 5/18/2026

Context, Reasoning, and Hierarchy: A Cost-Performance Study of Compound LLM Agent Design in an Adversarial POMDP

Cost-performance study of compound LLM agent design dimensions (context, reasoning, hierarchy) in adversarial environments.

Ax Tahreem Yasir, Wenbo Li, Sam Gilson, Sutapa Dey Tithi, Xiaoyi Tian, Tiffany Barnes 5/18/2026

Confirming Correct, Missing the Rest: LLM Tutoring Agents Struggle Where Feedback Matters Most

Benchmark evaluating 7 LLM tutoring agents on diagnostic precision for student solutions. Shows agents struggle with suboptimal feedback.

Ax Xavier Theimer-Lienhard, Mushtaha El-Amin, Fay Elhassan, Sahaj Vaidya, Victor Cartier-Negadi, David Sasu, Lars Klein, Mary-Anne Hartley 5/18/2026

Fully Open Meditron: An Auditable Pipeline for Clinical LLMs

Fully Open Meditron: Clinical LLM with complete transparent training pipeline including data provenance and curation procedures.

Ax Igor Bogdanov, Chung-Horng Lung, Thomas Kunz, Jie Gao, Adrian Taylor, Marzia Zaman 5/18/2026

FORGE: Self-Evolving Agent Memory With No Weight Updates via Population Broadcast

FORGE: Population-based protocol for evolving natural-language memory in ReAct LLM agents through prompt injection and reflexion loops.

Ax Sarah Martinson, Michael P. Brenner, Martyna Plomecka, Brian P. Williams, Nicholas G. Reich, Zahra Shamsi 5/18/2026

Prospective multi-pathogen disease forecasting using autonomous LLM-guided tree search

Autonomous LLM-guided tree search system for iterative disease forecasting model generation without manual expert curation.

Ax Can Zhou, Yulong Gao, Pian Yu 5/18/2026

Ensuring Logic in the Fog: Sound POMDP Synthesis with LTL Objectives

Sound POMDP synthesis with LTL objectives for autonomous agents navigating uncertain environments with temporal constraints.

Ax Jinze Wang, Yangchen Zeng, Tiehua Zhang, Lu Zhang, Yuze Liu, Yongchao Liu, Xingjun Ma, Zhu Sun 5/18/2026

Agent4POI: Agentic Context-Conditioned Affordance Reasoning for Multimodal Point-of-Interest Recommendation

Agent4POI: Agentic framework generating context-conditioned multimodal representations for POI recommendation at inference time.

Ax Dzung Pham, Kleomenis Katevas, Ali Shahin Shamsabadi, Hamed Haddadi 5/18/2026

AgentStop: Terminating Local AI Agents Early to Save Energy in Consumer Devices

AgentStop: Energy-efficient early termination strategy for local LLM-based agents on consumer devices.

Ax Plawan Kumar Rath, Rahul Maliakkal 5/18/2026

Quantization Undoes Alignment: Bias Emergence in Compressed LLMs Across Models and Precision Levels

Empirical study showing quantization induces bias emergence in LLMs across model families and precision levels, harming alignment.

Ax Yibin Wang, Yanjie Yang, Grace Melo Guerrero, Rodolfo M. Nayga Jr., Azlan Zahid 5/18/2026

An LLM-RAG Approach for Healthy Eating Index-Informed Personalized Food Recommendations

RAG-based food recommendation system using Healthy Eating Index and LLMs for personalized diet suggestions.

Ax Michael Y. Hu, Apurva Gandhi, Kyunghyun Cho, Tal Linzen, Pratyusha Sharma 5/18/2026

Always Learning, Always Mixing: Efficient and Simple Data Mixing All The Time

Universal data mixing method for language model training across pretraining, continual learning, and adaptation phases.

Ax Yoichi Ishibashi, Taro Yano, Masafumi Oyamada 5/18/2026

Effective Harness Engineering for Algorithm Discovery with Coding Agents

Studies effective harness design for LLM-based algorithm discovery, analyzing trade-offs in token budget allocation for evolutionary search.

Ax Nenad Petrovic, Andre Schamschurko, Yingjie Xu, Alois Knoll 5/18/2026

GenAI-Driven Approach to RISC-V Supply Chain Exploration

LLM and Vision-Language Model workflow for RISC-V supply chain analysis integrating multimodal data extraction and model-driven engineering.

Ax Bogdan Banu 5/18/2026

Do Biological Structural Guarantees Earn Their Complexity?

Empirical evaluation of biologically-inspired AI agent frameworks against simpler baselines across three deep benchmarks.

Ax Qingyun Zou, Feng Yu, Hongshi Tan, Bingsheng He, WengFai Wong 5/18/2026

Is Agentic AI Ready for Real-World Hardware Engineering? A Deep Dive with Phoenix-bench

Phoenix-bench: Benchmark of 511 hardware engineering tasks testing whether software-focused agentic AI transfers to realistic EDA workflows.

Ax Lucas Jing, Xinqi Wang, Liao Zhang, Simon S. Du 5/18/2026

PBT-Bench: Benchmarking AI Agents on Property-Based Testing

PBT-Bench: Benchmark of 100 property-based testing tasks evaluating AI agents on deriving semantic invariants from documentation.

Ax Abinand Nallathambi, Christopher Knight, Shantanu Ganguly, Wilfried Haensch, Anand Raghunathan 5/18/2026

A3D: Agentic AI flow for autonomous Accelerator Design

A3D: Agentic AI system for autonomous hardware accelerator design using LLMs and tool integration to reduce design labor.

Ax Alexander Du, Jianjun Ou, Danyang Zhuo, Matthew Lentz 5/18/2026

Hydra: Efficient, Correct Code Generation via Checkpoint-and-Rollback Support

Hydra: LLM code generation system using checkpoint-and-rollback for efficient error detection and repair during decoding.

Ax Fanxu Meng 5/18/2026

GQLA: Group-Query Latent Attention for Hardware-Adaptive Large Language Model Decoding

Group-Query Latent Attention mechanism for efficient LLM decoding across diverse hardware, improving upon DeepSeek-V2/V3 architecture.

Ax Vinil Pasupuleti, Siva Rama Krishna Varma Bayyavarapu, Shrey Tyagi 5/18/2026

Autonomous Intelligent Agents for Natural-Language-Driven Web Execution with Integrated Security Assurance

AI-driven autonomous testing framework using LLMs for web test automation with integrated security, addressing test suite maintenance failures.

Ax Kyle R. Chickering, Huijuan Wang, Mengxi Wu, Alexander Moreno, Muhao Chen, Xuezhe Ma, Daria Soboleva, Joel Hestness, Zhengzhong Liu, Eric Xing 5/18/2026

GQA-{\mu}P: The maximal parameterization update for grouped query attention

Maximal parameterization update rule for grouped query attention enabling hyperparameter transfer across LLM architectures.

Ax Milind Pandurang Jagre, Jia Huang, Dayvid V. R. Oliveira, Zhinan Cheng, Babak Seyed Aghazadeh, Puja Das, Chris Alvino, Jinda Han, Kailash Thiyagarajan 5/18/2026

Fortress: A Case Study in Stabilizing Search Recommendations via Temporal Data Augmentation and Feature Pruning

Framework for stabilizing recommendation system predictions through temporal data augmentation and feature pruning.

Ax Yihong Dong, Jiaru Qian, Haoran Zhang, Peixu Wang, Binhua Li, Zhi Jin, Yongbin Li, Ge Li, Xiaokang Yang, Xue Jiang 5/18/2026

From I/O to Code with Discovery Agent

Discovery agent system for automatic program synthesis from input-output behavior using LLMs.

Ax Sidharth Pulipaka, Stanislau Hlebik, Leonidas Raghav, Sahar Abdelnabi, Vyas Raina, Ivaxi Sheth, Mario Fritz 5/18/2026

Hidden in Memory: Sleeper Memory Poisoning in LLM Agents

Security analysis of sleeper memory poisoning attacks on LLM agents with persistent external memory.

Ax Marilyn Zhang, Tianfeng Chen, Fabi\'an Barzuna, Ankita Rathod, Mark E. Whiting 5/18/2026

LEAP: Trajectory-Level Evaluation of LLMs in Iterative Scientific Design

Trajectory-level evaluation framework for LLMs in iterative scientific design and autonomous laboratories.

Ax Ramon Vi\~nas Torn\'e, S\'ilvia F\`abregas Salazar, Soyon Park, Ivo Alexander Ban, Artyom Gadetsky, Nikita Doikov, Maria Brbi\'c 5/18/2026

PACER: Acyclic Causal Discovery from Large-Scale Interventional Data

Causal discovery method for DAGs from large-scale interventional data with improved scalability.

Ax Songwei Dong, Zihan Chen, Chengshuai Shi, Peng Wang, Jundong Li, Cong Shen 5/18/2026

Is One Score Enough? Rethinking the Evaluation of Sequentially Evolving LLM Memory

Diagnostic evaluation framework for LLM sequential memory, capturing forgetting and transfer patterns.

Ax Biswa Sengupta 5/18/2026

Representation Without Reward: A JEPA Audit for LLM Fine-Tuning

Study of JEPA principles applied to autoregressive LLM fine-tuning and latent representation learning.

Ax Lingzhe Zhang, Tong Jia, Yunpeng Zhai, Zixuan Xie, Chiming Duan, Minghua He, Philip S. Yu, Ying Li 5/18/2026

From Feedback Loops to Policy Updates: Reinforcement Fine-Tuning for LLM-Based Alpha Factor Discovery

Reinforcement fine-tuning approach for LLM-based discovery of alpha factors in quantitative trading.

Ax Gaojie Jin, Yong Tao, Lijia Yu, Tianjin Huang 5/18/2026

Margin-Adaptive Confidence Ranking for Reliable LLM Judgement

Framework for improving LLM confidence estimation and reliability in judgement tasks through margin-adaptive ranking.

Ax Jake Fawkes, Jason Hartford 5/18/2026

$f$-Trajectory Balance: A Loss Family for Tuning GFlowNets, Generative Models, and LLMs with Off- and On-Policy Data

Loss function family for training GFlowNets, generative models, and LLMs using off- and on-policy data.

Ax Shubhi Asthana, Bing Zhang, Chad DeLuca, Hima Patel, Ruchi Mahindru 5/18/2026

Runtime-Structured Task Decomposition for Agentic Coding Systems

Runtime-structured task decomposition for agentic LLM coding systems improving debuggability and reducing retry costs in software engineering.

Ax Yikun Zhang, Xiwei Cheng, Tianyu Liu, Yuanqi Du, Wengong Jin 5/18/2026

DrugSAGE:Self-evolving Agent Experience for Efficient State-of-the-Art Drug Discovery

DrugSAGE: Self-evolving agent that accumulates and reuses experience for efficient AutoML-style drug discovery model building.

Ax Shangjian Yin, Yu Fu, Yue Dong, Zhouxing Shi 5/18/2026

GRLO: Towards Generalizable Reinforcement Learning in Open-Ended Environments from Zero

GRLO framework combining RLHF and RLVR paradigms for generalizable LLM post-training across open-ended environments.

Ax A H M Rezaul Karim, Ozlem Uzuner 5/18/2026

Retrieval-Augmented Large Language Models for Schema-Constrained Clinical Information Extraction

Retrieval-augmented LLMs extract structured clinical information from nurse-patient transcripts for medical documentation automation.

Ax Swayamjit Saha, Subhabrata Das, Haonan Duan, Xiao-Yang Liu 5/18/2026

Hybrid LLM-based Intelligent Framework for Robot Task Scheduling

LLM-based framework for robot task scheduling using natural language processing to optimize time efficiency and resource allocation.

Ax Vincent-Daniel Yun, Junhyuk Jo, Sai Praneeth Karimireddy, Sunwoo Lee 5/18/2026

Ghosted Layers: Unconstrained Activation Alignment for Recovering Layer-Pruned LLMs

Ghosted Layers recovers pruned LLM performance by solving boundary activation alignment without retraining pruned decoder blocks.

Ax Yufeng Du, Phillip Harris, Minyang Tian, Eliu A Huerta, Srikanth Ronanki, Subendhu Rongali, Aram Galstyan, Hao Peng 5/18/2026

RoPE Distinguishes Neither Positions Nor Tokens in Long Contexts, Provably

Theoretical analysis proving RoPE positional embeddings lose locality bias and position discrimination in long-context Transformers.

Ax Xian Gao, Bo Hui, Min-Te Sun, Wei-Shinn Ku 5/18/2026

On the Fragility of Data Attribution When Learning Is Distributed

Demonstrates data attribution values can be manipulated in distributed training while preserving model utility, affecting ML governance.

Ax Jinyuan Li, Langlin Huang, Chengsong Huang, Shaoyang Xu, Donghong Cai, Yuyi Yang, Wenxuan Zhang, Jiaxin Huang 5/18/2026

Process Rewards with Learned Reliability

BetaPRM: Process Reward Model predicting step-level success probability and reliability scores for LLM reasoning verification.

Ax Jaehun Jung, Hyunwoo Kim, Brandon Cui, Ximing Lu, David Acuna, Prithviraj Ammanabrolu, Yejin Choi 5/18/2026

DeltaPrompts: Escaping the Zero-Delta Trap in Multimodal Distillation

DeltaPrompts improves multimodal distillation by identifying and replacing ineffective prompts that don't change student model outputs.

Ax Haizhong Zheng, Yizhuo Di, Jiahui Wang, Shuowei Jin, Xueshen Liu, Yongji Wu, Z. Morley Mao, Ion Stoica, Jiawei Zhao, Beidi Chen 5/18/2026

AstraFlow: Dataflow-Oriented Reinforcement Learning for Agentic LLMs

AstraFlow: dataflow-oriented reinforcement learning system for scaling agentic LLMs with multi-policy training on heterogeneous, elastic compute resources.

Ax Penghui Li, Hong Yau Chong, Yinzhi Cao, Junfeng Yang 5/18/2026

Detecting Privilege Escalation in Polyglot Microservices via Agentic Program Analysis

Agentic program analysis system for detecting privilege escalation vulnerabilities in polyglot microservices through cross-service interaction modeling.

Ax Hojun Chung, Junseo Lee, Songhwai Oh 5/18/2026

Offline Reinforcement Learning with Universal Horizon Models

Offline reinforcement learning method using universal horizon models to reduce compounding errors from repeated self-generated state inference.

Ax Shaoke Xi, ChonLam Lao, Boyi Jia, Jiaqi Gao, Zhipeng Zhang, Jiamin Cao, Brian Sutioso, Erci Xu, Minlan Yu, Kui Ren, Yong Li, Zhengping Qian, Ennan Zhai, Jingren Zhou 5/18/2026

A Few GPUs, A Whole Lotta Scale: Faithful LLM Training Emulation with PrismLLM

PrismLLM: tool for emulating large-scale LLM training on small GPU clusters to enable debugging and optimization without accessing production infrastructure.

Ax Ali J Alrasheed, Aryan Yazdan Parast, Basim Azam, James Bailey, Naveed Akhtar 5/18/2026

Latent Video Prediction Learns Better World Models

Systematic evaluation of latent video prediction models as world models, analyzing robustness of V-JEPA, VideoPrism, and VideoMAEv2 across five axes.

Ax Hyesoo Hong, Minsoo Kim, Wonje Jeung, Sangyeon Yoon, Dongjae Jeon, Albert No 5/18/2026

VLMs Trace Without Tracking: Diagnosing Failures in Visual Path Following

Study of line-tracing ability in vision-language models, diagnosing failures in basic visual path-following operations across controlled tasks.

Ax Jaeseung Heo, Kyeongheung Yun, Youngbin Choi, Sehyun Hwang, Jungseul Ok, Dongwoo Kim 5/18/2026

Interaction-Aware Influence Functions for Group Attribution

Interaction-aware influence functions measuring how groups of training examples jointly affect model predictions, capturing redundancy and complementarity.