Ax Hongbo Jin, Rongpeng Zhu, Zhongjing Du, Xu Jiang, Jingqi Tian, Qiaoman Zhang, Jiayu Ding 5/7/2026

DGPO: Distribution Guided Policy Optimization for Fine Grained Credit Assignment

DGPO algorithm for fine-grained credit assignment in RL-based LLM alignment, improving reasoning step isolation in chain-of-thought reasoning.

Ax Ahmadreza Eslaminia, Chuhan Cai, Cameron Smith, Ruo-Syuan Mei, Shichen Li, Rajiv Malhotra, Klara Nahrstedt, Chenhui Shao 5/7/2026

LLM-ADAM: A Generalizable LLM Agent Framework for Pre-Print Anomaly Detection in Additive Manufacturing

LLM-based agent framework for detecting anomalies in additive manufacturing 3D printing processes.

Ax Negar Arabzadeh, Wenjie Ma, Sewon Min, Matei Zaharia 5/7/2026

RAG over Thinking Traces Can Improve Reasoning Tasks

RAG applied to reasoning tasks by retrieving intermediate thinking traces instead of documents, improving math and code generation performance.

Ax Hahyeon Choi, Nojun Kwak 5/7/2026

Toward Structural Multimodal Representations: Specialization, Selection, and Sparsification via Mixture-of-Experts

S3 framework for multimodal learning using mixture-of-experts to specialize, select, and route semantic experts for task-specific needs.

Ax JF Bastien, Sam D'Amico 5/7/2026

VLMaxxing through FrameMogging Training-Free Anti-Recomputation for Video Vision-Language Models

Training-free optimization for video vision-language models that reuses stable visual state to reduce redundant computation.

Ax Lina Zhang, Tonmoy Monsoor, Mehmet Efe Lorasdagi, Prateik Sinha, Chong Han, Peizheng Li, Yuan Wang, Jessica Pasqua, Colin McCrimmon, Rajarshi Mazumder, Vwani Roychowdhury 5/7/2026

Can Multimodal Large Language Models Understand Pathologic Movements? A Pilot Study on Seizure Semiology

Pilot study evaluating multimodal LLMs for zero-shot recognition of pathological seizure movements in neurological disorder videos.

Ax Yipeng Ouyang, Yi Xiao, Yuhao Gu, Xianwei Zhang 5/7/2026

SkCC: Portable and Secure Skill Compilation for Cross-Framework LLM Agents

SkCC: portable skill compiler for LLM agents addressing 40% performance variation across frameworks by converting SKILL.md specs to platform-specific formats.

Ax Xiao Zhang, Yafei Li, Ruixiang Wang, Wei Wei, Shuo He, Mingliang Xu 5/7/2026

Local Truncation Error-Guided Neural ODEs for Large Scale Traffic Forecasting

Neural ODE method for large-scale traffic forecasting handling continuous dynamics and discrete anomalies with truncation error guidance.

Ax Akshat Singh Jaswal, Ashish Baghel, Paras Chopra 5/7/2026

Discovering Reinforcement Learning Interfaces with Large Language Models

LLMs synthesize complete RL task interfaces (observations and reward functions) automating environment interface design for new tasks.

Ax Doojin Baek, Gyubin Lee, Junyeob Baek, Hosung Lee, Sungjin Ahn 5/7/2026

Learning to Theorize the World from Observation

Theory-building world models inspired by developmental cognition: learning internal representations of world dynamics from observation.

Ax Al Zadid Sultan Bin Habib, Gianfranco Doretto, Donald A. Adjeroh 5/7/2026

DynaTab: Dynamic Feature Ordering as Neural Rewiring for High-Dimensional Tabular Data

DynaTab: dynamic feature ordering architecture for high-dimensional tabular data using neural rewiring with permutation complexity detection.

Ax Haoyu Zhang, Mohammad Zandsalimy, Shanu Sushmita 5/7/2026

Exposing LLM Safety Gaps Through Mathematical Encoding:New Attacks and Systematic Analysis

Analysis of LLM safety bypasses using mathematical encoding (set theory, logic, quantum mechanics) achieving 46-56% attack success across models.

Ax Zhenghao Huang, Huilin Yao, Kaikai Wang, Lin Shu 5/7/2026

Learning Generalizable Action Representations via Pre-training AEMG

AEMG: self-supervised pre-training framework for learning generalizable action representations from electromyography across subjects and devices.

Ax Thanh Dat Hoang, Thanh Trung Huynh, Matthias Weidlich, Thanh Tam Nguyen, Tong Chen, Hongzhi Yin, Quoc Viet Hung Nguyen 5/7/2026

FINER-SQL: Boosting Small Language Models for Text-to-SQL

FINER-SQL boosts small language models for text-to-SQL tasks, enabling efficient on-premise deployment with improved reasoning and instruction following.

Ax Tianze Han, Beining Xu, Hanbo Zhang, Yongming Lu 5/7/2026

Detecting Stealth Sycophancy in Mental-Health Dialogue with Dynamic Emotional Signature Graphs

Method for detecting sycophancy in conversational AI therapists using dynamic emotional signature graphs without LLM judges.

Ax Severin Ye, Xiao Kong, Xiaopeng He, Guangsu Yan, Dongsuk Oh 5/7/2026

CuraView: A Multi-Agent Framework for Medical Hallucination Detection with GraphRAG-Enhanced Knowledge Verification

CuraView multi-agent framework using GraphRAG for detecting hallucinations in LLM-generated medical discharge summaries from EHRs.

Ax Ishrith Gowda (University of California, Berkeley) 5/7/2026

MEMSAD: Gradient-Coupled Anomaly Detection for Memory Poisoning in Retrieval-Augmented Agents

MEMSAD: security framework for detecting gradient-based anomalies and memory poisoning attacks on retrieval-augmented LLM agents via Stackelberg game analysis.

Ax Kangkang Wang, Qinting Jiang, Wanping Zhang, Bowen Ren, Shengzhao Wen 5/7/2026

MHPR: Multidimensional Human Perception and Reasoning Benchmark for Large Vision-Languate Models

MHPR benchmark for evaluating vision-language models on human perception and reasoning tasks across individual, multi-person, and interaction dimensions.

Ax Zhongjian Zhang, Yue Yu, Mengmei Zhang, Junping Du, Xiao Wang, Chuan Shi 5/7/2026

Revisiting Graph-Tokenizing Large Language Models: A Systematic Evaluation of Graph Token Understanding

Systematic evaluation challenging whether graph-tokenizing LLMs effectively understand graph structure or merely process tokens.

Ax Eric H. C. Chow 5/7/2026

A Skill-Based AI Agentic Pipeline for Library of Congress Subject Indexing

Modular AI agent pipeline using discrete skills for automating Library of Congress subject indexing in library cataloging.

Ax John Yang, Kilian Lieret, Jeffrey Ma, Parth Thakkar, Dmitrii Pedchenko, Sten Sootla, Emily McMilin, Pengcheng Yin, Rui Hou, Gabriel Synnaeve, Diyi Yang, Ofir Press 5/7/2026

ProgramBench: Can Language Models Rebuild Programs From Scratch?

Benchmark testing language model agents' ability to build complete software projects from scratch with minimal human oversight.

Ax JuneHyoung Kwon, JungMin Yun, YoungBin Kim 5/7/2026

Erase Persona, Forget Lore: Benchmarking Multimodal Copyright Unlearning in Large Vision Language Models

Benchmark for evaluating machine unlearning in vision-language models to remove memorized copyrighted visual content.

Ax Jorge L. Ruiz Williams 5/7/2026

HeadQ: Model-Visible Distortion and Score-Space Correction for KV-Cache Quantization

KV-cache quantization method measuring error in model-visible coordinates for efficient LLM inference with low-rank residuals.

Ax Maxim Chupilkin 5/7/2026

Multi-Agent Strategic Games with LLMs

Empirical study of LLMs as agents in repeated games, testing strategic behavior from international relations theory.

Ax Jiaxi Li, Lu Yin, Li Shen, Jinjin Xu, Yuhui Liu, Wenwu Wang, Shiwei Liu, Xilu Wang 5/7/2026

ELAS: Efficient Pre-Training of Low-Rank Large Language Models via 2:4 Activation Sparsity

ELAS enables efficient LLM pre-training via low-rank adaptation combined with 2:4 activation sparsity for GPU acceleration.

Ax Timon Homberger, Finn Lukas Busch, Jes\'us Gerardo Ortega Peimbert, Quantao Yang, Olov Andersson 5/7/2026

FUS3DMaps: Scalable and Accurate Open-Vocabulary Semantic Mapping by 3D Fusion of Voxel- and Instance-Level Layers

Open-vocabulary semantic mapping for robots using 3D fusion of voxel and instance-level semantic embeddings.

Ax Xing Zhang, Keyu Zhang, Taohong Zhu, Anbang Ruan 5/7/2026

Tailored Prompts, Targeted Protection: Vulnerability-Specific LLM Analysis for Smart Contracts

LLM-based framework for smart contract vulnerability detection with vulnerability-specific prompts and 31K dataset.

Ax Zhifeng Hao, Zhongjie Chen, Junhao Lu, Shengyin Yu, Guimin Hu, Keli Zhang, Ruichu Cai, Boyan Xu 5/7/2026

SERE: Structural Example Retrieval for Enhancing LLMs in Event Causality Identification

SERE method uses structural example retrieval to enhance LLM performance on event causality identification and reduce causal hallucination.

Ax Ruichu Cai, Juntao Gan, Miao Mai, Zhifeng Hao, Boyan Xu 5/7/2026

SAM-NER: Semantic Archetype Mediation for Zero-Shot Named Entity Recognition

SAM-NER framework for zero-shot NER using semantic archetype mediation to handle domain and schema shifts in LLMs.

Ax Mengchu Li, Jin Zhu, Jinglai Li, Chengchun Shi 5/7/2026

Segmenting Human-LLM Co-authored Text via Change Point Detection

Algorithms for segmenting human-written and LLM-generated text in co-authored documents using change point detection.

Ax Juneyoung Park 5/7/2026

Rethinking the Rank Threshold for LoRA Fine-Tuning

Analysis of LoRA rank threshold requirements for fine-tuning, examining theoretical conditions and practical implications for cross-entropy loss.

Ax Sergio Mendoza, Cedric Bhihe, Natalia Zamora, David Modesto, Jose Martin Bugallo Batalla, Jesus Gomez Canovas, Rafel Palomo Avellaneda, Miguel Perez Espinosa 5/7/2026

A Workflow-Oriented Framework for Asynchronous Human-AI Collaboration in Hybrid and Compute-Intensive HPC Environments

Workflow framework for asynchronous human-AI collaboration in HPC environments with checkpoint-based intervention for high-stakes applications.

Ax JuneHyoung Kwon, MiHyeon Kim, Eunju Lee, JungMin Yun, Byeonggeuk Lim, YoungBin Kim 5/7/2026

Before Forgetting, Learn to Remember: Revisiting Foundational Learning Failures in LVLM Unlearning Benchmarks

Identifies under-memorization failures in LVLM unlearning benchmarks that use fictitious identities for privacy evaluation.

Ax Carla Soares, Gabriel Moreira, Ana Paula Camargo, Fabio Henrique Scacabarozi, Nicole Davila, Marselle Silva 5/7/2026

AI Advocate: Educational Path to Transform Squads to the Future

Case study on upskilling software teams with AI Advocates to enable human-AI collaborative development.

Ax Hao Wu, Yuqi Li, Yuan Gao, Fan Xu, Fan Zhang, Kun Wang, Penghao Zhao, Qiufeng Wang, Yizhou Zhao, Weiyan Wang, Yingli Tian, Xian Wu, Xiaomeng Huang 5/7/2026

RoboAlign-R1: Distilled Multimodal Reward Alignment for Robot Video World Models

RoboAlign-R1 uses reward alignment to improve robot video world models for instruction following and manipulation tasks.

Ax Serhii Zabolotnii 5/7/2026

TRACE: A Metrologically-Grounded Engineering Framework for Trustworthy Agentic AI Systems in Operationally Critical Domains

TRACE framework for trustworthy agentic AI systems in critical domains, combining architecture, metrics, and human supervision protocols.

Ax Jaeyun Lee, Junyoung Koh, Zeynel Tok, Hunar Batra, Ronald Clark 5/7/2026

MCJudgeBench: A Benchmark for Constraint-Level Judge Evaluation in Multi-Constraint Instruction Following

MCJudgeBench benchmark evaluates LLM judges on constraint-level accuracy in multi-constraint instruction following tasks.

Ax Qichao Wang, Yunhong Lu, Hengyuan Cao, Junyi Zhang, Min Zhang 5/7/2026

DMGD: Train-Free Dataset Distillation with Semantic-Distribution Matching in Diffusion Models

Train-free dataset distillation using semantic-distribution matching in diffusion models without fine-tuning.

Ax Zhihan Jiang, Mengyuan Millie Wu, Ruishi Zou, Shiyu Xu, Xun Qian, Emma Macmanus, Steven Liao, Ping Zhang, Bingsheng Yao, Tingyu Cheng, James L. David, Nabila El-Bassel, Lena Mamykina, Frances R. Levin, Ryan Sultan, Dakuo Wang, Xuhai Xu 5/7/2026

Deco: Extending Personal Physical Objects into Pervasive AI Companion through a Dual-Embodiment Framework

Deco dual-embodiment framework extending emotional bonds from physical objects to AI companion agents.

Ax Geert Heyman, Frederik Vandeputte 5/7/2026

Steer Like the LLM: Activation Steering that Mimics Prompting

Framework to improve activation steering in LLMs by distilling prompt-based steering behavior into interpretable models.

Ax Lisa C. Adams, Linus Marx, Erik Thiele Orberg, Keno Bressem, Sebastian Ziegelmayer, Denise Bernhardt, Markus Graf, Marcus R. Makowski, Stephanie E. Combs, Florian Matthes, Jan C. Peeken 5/7/2026

Atomic Fact-Checking Increases Clinician Trust in Large Language Model Recommendations for Oncology Decision Support: A Randomized Controlled Trial

Randomized trial showing atomic fact-checking increases clinician trust in LLM oncology decision support recommendations.

Ax Daniel Drucker, Kyle Mahowald 5/7/2026

The Counterexample Game: Iterated Conceptual Analysis and Repair in Language Models

Language models perform iterative conceptual analysis through counterexample generation and definition repair chains.

Ax Jianjie Fang, Yingshan Lei, Qin Wan, Ziyou Wang, Yuchao Huang, Yongyan Xu, Baining Zhao, Weichen Zhang, Chen Gao, Xinlei Chen, Yong Li 5/7/2026

iWorld-Bench: A Benchmark for Interactive World Models with a Unified Action Generation Framework

iWorld-Bench comprehensive benchmark for training and evaluating interactive world models with unified action generation framework.

Ax Stanislav Kirpichenko, Andrei Konstantinov, Lev Utkin 5/7/2026

TabSurv: Adapting Modern Tabular Neural Networks to Survival Analysis

TabSurv adapts modern tabular neural networks to survival analysis using Weibull and non-parametric methods.

Ax Jonathan Steinberg, Oren Gal 5/7/2026

MOSAIC-Bench: Measuring Compositional Vulnerability Induction in Coding Agents

MOSAIC-Bench benchmark measuring safety vulnerabilities in coding agents through compositional task decomposition.

Ax Danny Hoang, Ryan Matthiessen, Christopher Miller, Nasir Mannan, Ruby ElKharboutly, David Gorsich, Matthew P. Castanier, Farhad Imani 5/7/2026

Physics-Grounded Multi-Agent Architecture for Traceable, Risk-Aware Human-AI Decision Support in Manufacturing

MAKA: Multi-agent human-in-the-loop system for CNC aerospace manufacturing combining LLM text generation with risk-constrained numerical workflows and auditable provenance.

Ax Sebastian Wind, Tri-Thien Nguyen, Jeta Sopa, Mahshad Lotfinia, Sebastian Bickelhaup, Michael Uder, Harald K\"ostler, Gerhard Wellein, Sven Nebelung, Daniel Truhn, Andreas Maier, Soroosh Tayebi Arasteh 5/7/2026

Safety and accuracy follow different scaling laws in clinical large language models

SaFE-Scale framework showing clinical LLM safety and accuracy follow different scaling laws; safety doesn't necessarily improve with model size or context.

Ax Peter Slattery, Alexander K. Saeri, Emily A. C. Grundy, Jess Graham, Michael Noetel, Risto Uuk, James Dao, Soroush Pour, Stephen Casper, Neil Thompson 5/7/2026

The AI risk repository: A meta-review, database, and taxonomy of risks from artificial intelligence

Meta-review database and taxonomy of AI risks including terminology standardization across research, policymakers, and companies for shared risk discussion.

Ax Ivaxi Sheth, Jan Wehner, Sahar Abdelnabi, Ruta Binkyte, Mario Fritz 5/7/2026

Safety Must Precede the Deployment of Open-Ended AI

Position paper on safety requirements for open-ended AI agents with autonomous behavior generation and self-evolution capabilities before deployment.

Ax Zhiyuan Li, Wenshuai Zhao, Joni Pajarinen 5/7/2026

Closed-Loop Vision-Language Planning for Multi-Agent Coordination

COMPASS: Multi-agent framework integrating vision-language models for closed-loop planning in cooperative multi-agent RL, improving sample efficiency and generalization.