Ax Guangyu Feng, Huanzhi Mao, Prabal Dutta, Joseph E. Gonzalez 5/15/2026

Concurrency without Model Changes: Future-based Asynchronous Function Calling for LLMs

AsyncFC: Execution framework enabling concurrent function calling for LLM agents by decoupling decoding from function execution, reducing latency.

Ax Ziyin Zhang, Zihan Liao, Hang Yu, Peng Di, Rui Wang 5/15/2026

ML-Embed: Inclusive and Efficient Embeddings for a Multilingual World

ML-Embed: Suite of inclusive multilingual text embedding models addressing computational costs, linguistic coverage, and transparency barriers in open-source framework.

Ax Daniel Asare Kyei, Alimatu Saadia-Yussiff, Maame G. Asante-Mensah, Abdul Lateef-Yussiff, Charles Roland Haruna, Derry Emmanuel 5/15/2026

Novel Dynamic Batch-Sensitive Adam Optimiser for Vehicular Accident Injury Severity Prediction

DBS-Adam optimizer for deep learning on imbalanced/sequential datasets, applied to vehicular accident injury severity prediction.

Ax Renning Pang, Tian Lan, Leyuan Liu, Xiaoming Huang, Piao Tong, Xiaosong Zhang 5/15/2026

Improving Multi-turn Dialogue Consistency with Self-Recall Thinking

Method improving LLM multi-turn dialogue consistency using self-recall thinking to track non-adjacent turn dependencies and reduce context bottlenecks.

Ax Connor Douglas, Joel Persson, Foster Provost 5/15/2026

Logging Policy Design for Off-Policy Evaluation

Research on designing logging policies to minimize off-policy evaluation error for treatment policies like recommender systems.

Ax Sining Ang, Yuguang Yang, Canyu Chen, Yan Wang 5/15/2026

CLOVER: Closed-Loop Value Estimation \& Ranking for End-to-End Autonomous Driving Planning

CLOVER: Closed-loop value estimation and ranking for end-to-end autonomous driving planning with training-evaluation mismatch resolution.

Ax Laleh Nourian, Anisa Callis, Stephanie Patterson, Jadeline Miao, Jamison Heard, Garreth W. Tigwell 5/15/2026

Understanding How International Students in the U.S. Are Using Conversational AI to Support Cross-Cultural Adaptation

Study on international students using conversational AI (ChatGPT, Gemini) for cross-cultural adaptation support.

Ax Xiaohua Zhan, Kazuki Egashira, Robin Staab, Mark Vero, Martin Vechev 5/15/2026

Widening the Gap: Exploiting LLM Quantization via Outlier Injection

Security research on adversarial attacks exploiting LLM quantization via outlier injection, showing quantized models can exhibit malicious behavior.

Ax Yi Zhang, Yinda Chen, Che Liu, Zeyuan Ding, Jin Xu, Shilong Zou, Junwei Liao, Jiayu Hu, Xiancong Ren, Xiaopeng Zhang, Yechi Liu, Haoyuan Shi, Zecong Tang, Haosong Sun, Renwen Cui, Kuishu Wu, Wenhai Liu, Yang Xu, Yingji Zhang, Yidong Wang, Senkang Hu, Jinpeng Lu, Nga Teng Chan, Yechen Wu, Yong Dai, Jian Tang, Xiaozhu Ju 5/15/2026

Pelican-Unified 1.0: A Unified Embodied Intelligence Model for Understanding, Reasoning, Imagination and Action

Pelican-Unified 1.0: Embodied foundation model using single VLM for unified understanding, reasoning, imagination, and action in shared semantic space.

Ax Zhengxi Lu, Zhiyuan Yao, Zhuowen Han, Zi-Han Wang, Jinyang Wu, Qi Gu, Xunliang Cai, Weiming Lu, Jun Xiao, Yueting Zhuang, Yongliang Shen 5/15/2026

Self-Distilled Agentic Reinforcement Learning

On-Policy Self-Distillation (OPSD) method for RL-based LLM agents using dense token-level supervision from teacher branch with privileged context for multi-turn interactions.

Ax Ryan Wei Heng Quek, Sanghyuk Lee, Alfred Wei Lun Leong, Arun Verma, Alok Prakash, Nancy F. Chen, Bryan Kian Hsiang Low, Daniela Rus, Armando Solar-Lezama 5/15/2026

MeMo: Memory as a Model

MeMo framework encodes new knowledge into dedicated memory model while keeping LLM frozen, enabling efficient knowledge incorporation for domain-specific applications.

Ax Pratinav Seth, Vinay Kumar Sankarapu 5/15/2026

Position: Behavioural Assurance Cannot Verify the Safety Claims Governance Now Demands

Position paper arguing behavioral assurance cannot verify safety claims demanded by AI governance frameworks (2019-2026).

Ax Shashwat Goel, Nikhil Chandak, Arvindh Arun, Ameya Prabhu, Steffen Staab, Moritz Hardt, Maksym Andriushchenko, Jonas Geiping 5/15/2026

FutureSim: Replaying World Events to Evaluate Adaptive Agents

FutureSim grounded simulation replaying real-world events chronologically to evaluate adaptive agents' ability to forecast beyond knowledge cutoff.

Ax Ziyu Guo, Rain Liu, Xinyan Chen, Pheng-Ann Heng 5/15/2026

ATLAS: Agentic or Latent Visual Reasoning? One Word is Enough for Both

ATLAS framework comparing agentic reasoning (code/tool calls) vs latent reasoning (learnable embeddings) for visual reasoning with efficiency trade-offs.

Ax Yumeng Fu, Jiayin Zhu, Lingling Zhang, Wenjun Wu, Bo Zhao, Shaoxuan Ma, Yushun Zhang, Jun Liu 5/15/2026

GeoLaux: A Benchmark for Evaluating MLLMs' Geometry Performance on Long-Step Problems Requiring Auxiliary Lines

GeoLaux benchmark with 2186 annotated geometry problems for evaluating MLLMs on long-step reasoning requiring auxiliary line construction.

Ax Zejun Li, Yingxiu Zhao, Jiwen Zhang, Siyuan Wang, Yang Yao, Runzhou Zhao, Jun Song, Bo Zheng, Zhongyu Wei 5/15/2026

Mixture-of-Visual-Thoughts: Exploring Context-Adaptive Reasoning Mode Selection for General Visual Reasoning

Mixture-of-Visual-Thoughts adaptive reasoning paradigm unifying multiple visual reasoning modes with context-based mode selection for general visual reasoning.

Ax Davi Bastos Costa, Renato Vicente 5/15/2026

Deceive, Detect, and Disclose: Large Language Models Play Mini-Mafia

Mini-Mafia game framework analyzing multi-agent LLM interactions and social deduction capabilities, providing theoretical understanding of collective agent outcomes.

Ax Yixu Wang, Xin Wang, Yang Yao, Xinyuan Li, Xibang Yang, Yan Teng, Xingjun Ma, Yingchun Wang 5/15/2026

AgenticEval: Toward Agentic and Self-Evolving Safety Evaluation of Large Language Models

AgenticEval paradigm for continuous self-evolving safety evaluation of LLMs, addressing dynamic risks and evolving regulations in high-stakes deployments.

Ax Wonjoong Kim, Sangwu Park, Yeonjun In, Sein Kim, Dongha Lee, Chanyoung Park 5/15/2026

Beyond the Final Answer: Evaluating the Reasoning Trajectories of Tool-Augmented Agents

TRACE framework for evaluating tool-augmented agent trajectories beyond final answers, assessing efficiency, hallucination, and adaptivity without ground-truth annotation.

Ax Mingyu Zhang, Lifeng Zhuo, Tianxi Tan, Guocan Xie, Xian Nie, Yan Li, Renjie Zhao, Zizhu He, Ziyu Wang, Jiting Cai, Yong-Lu Li 5/15/2026

IPR-1: Interactive Physical Reasoner

Interactive Physical Reasoner agent with G2U benchmark (1000+ games) for learning physics and causality through environment interaction with visual domain gaps.

Ax Wei Wu, Liyi Chen, Congxi Xiao, Tianfu Wang, Qimeng Wang, Chengqiang Lu, Yan Gao, Yi Wu, Yao Hu, Hui Xiong 5/15/2026

Anti-Length Shift: Dynamic Outlier Truncation for Training Efficient Reasoning Models

Dynamic outlier truncation method for training efficient reasoning models, addressing verbosity and deployment costs in RL-enhanced chain-of-thought systems.

Ax Kaixiang Wang, Yidan Lin, Jiong Lou, Zhaojiacheng Zhou, Bunyod Suvonov, Jie Li 5/15/2026

E-mem: Multi-agent based Episodic Context Reconstruction for LLM Agent Memory

Multi-agent episodic memory system for LLM agents using context reconstruction to enable System 2 reasoning with logical integrity over extended interactions.

Ax Mingda Zhang, Wenjin Liu, Tiesunlong Shen, Qika Lin, Rui Mao, Erik Cambria, Xiaoying Tang, Haoran Luo 5/15/2026

FlowSteer: Towards Agents Designing Agentic Workflows via Reinforced Progressive Canvas Editing

Agent that designs agentic workflows via reinforced canvas editing, addressing workflow construction, execution feedback, and error repair during long-horizon tasks.

Ax Yuanzhe Li, Jianing Deng, Jingtong Hu, Tianlong Chen, Song Wang, Huanrui Yang 5/15/2026

Dynamic Mixed-Precision Routing for Efficient Multi-step LLM Interaction

Dynamic mixed-precision routing system for multi-step LLM interactions reducing inference costs by adaptively routing tasks to smaller quantized models.

Ax Xi Wang, Anushri Suresh, Alvin Zhang, Rishi More, William Jurayj, Benjamin Van Durme, Mehrdad Farajtabar, Daniel Khashabi, Eric Nalisnick 5/15/2026

Conformal Thinking: Risk Control for Reasoning on a Compute Budget

Conformal inference approach for adaptive reasoning in LLMs with fixed compute budgets, balancing risk and accuracy through principled token allocation.

Ax Vlad Vinogradov, Alisa Vinogradova, Luba Greenwood, Ilya Yasny, Dmitry Kobyzev, Shoman Kasbekar, Kong Nguyen, Dmitrii Radkevich, Roman Doronin, Andrey Doronichev 5/15/2026

Hunt Globally: Wide Search AI Agents for Drug Asset Scouting in Investing, Business Development, and Competitive Intelligence

AI agent system for pharmaceutical competitive intelligence and drug asset discovery across global non-English channels and patent databases.

Ax Ayush Sawarni, Jiyuan Tan, Vasilis Syrgkanis 5/15/2026

CausalReasoningBenchmark: A Real-World Benchmark for Disentangled Evaluation of Causal Identification and Estimation

Benchmark disentangling causal identification from estimation in 173 queries across 132 datasets for evaluating automated causal inference systems.

Ax Borja Requena, Austin Letson, Krystian Nowakowski, Izan Beltran-Ferreiro, Leopoldo Sarra 5/15/2026

A Minimal Agent for Automated Theorem Proving

Minimal agentic baseline for automated theorem proving implementing iterative refinement, library search, and context management with comparative evaluation.

Ax Bingbing Wang, Jing Li, Ruifeng Xu 5/15/2026

Proactive Memory for Ad-Hoc Recall over Streaming Dialogues

STEM-Bench first benchmark for memory evaluation in streaming dialogue settings, addressing ad-hoc recall requirements over infinite conversation streams.

Ax Ziqin Gong, Ning Li, Huaikang Zhou 5/15/2026

LLMs learn scientific taste from institutional traces across the social sciences

Study on whether institutional publication traces can train LLMs to make evaluative judgments on untested ideas in low-verifiability domains.

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 5/15/2026

Graph of States: Solving Abductive Tasks with Large Language Models

Graph of States framework for solving abductive reasoning tasks with LLMs, addressing unstructured state representation issues in logical reasoning tasks.

Ax Manasa Bharadwaj, Yolanda Liu, InJung Yang, Sungil Kim, Nikhil Verma, KoKeun Kim, Kevin Ferreira, YoungJoon Kim 5/15/2026

PersonalHomeBench: Evaluating Agents in Personalized Smart Homes

PersonalHomeBench benchmark for evaluating foundation models as agentic assistants in personalized smart home environments with iteratively built household states.

Ax Yue Ma, Ziyuan Yang, Yi Zhang 5/15/2026

Catching the Infection Before It Spreads: Foresight-Guided Defense in Multi-Agent Systems

Defense framework against infectious jailbreaks in multi-agent systems using foresight-guided mechanisms to prevent compromise spread across collaborative agents.

Ax Xiao Jia 5/15/2026

NeuroState-Bench: A Human-Calibrated Benchmark for Commitment Integrity in LLM Agent Profiles

NeuroState-Bench human-calibrated benchmark evaluating whether LLM agent profiles preserve commitments across multi-turn tasks using side-query probes.

Ax Rahul Kumar 5/15/2026

The Compliance Trap: How Structural Constraints Degrade Frontier AI Metacognition Under Adversarial Pressure

SCHEMA evaluation framework measuring cognitive collapse and metacognitive degradation in frontier AI models under adversarial pressure and high-stakes scenarios.

Ax Zirui Tang, Xuanhe Zhou, Yumou Liu, Linchun Li, Yukai Wu, Weizheng Wang, Hongzhang Huang, Wei Zhou, Jun Zhou, Jiachen Song, Shaoli Yu, Jinqi Wang, Zihang Zhou, Hongyi Zhou, Yuting Lv, Jinyang Li, Jiashuo Liu, Ruoyu Chen, Chunwei Liu, GuoLiang Li, Jihua Kang, Fan Wu 5/15/2026

Workspace-Bench 1.0: Benchmarking AI Agents on Workspace Tasks with Large-Scale File Dependencies

Workspace-Bench 1.0 benchmark for evaluating AI agents on realistic workspace tasks with complex file dependencies and implicit/explicit relationships.

Ax Qiyao Liang, Risto Miikkulainen, Ila Fiete 5/15/2026

Attractor Geometry of Transformer Memory: From Conflict Arbitration to Confident Hallucination

Mechanistic study of language model failures examining conflict between parametric and working memory, and hallucination, using geometric analysis of attention patterns.

Ax Junwei Liao, Haoting Shi, Ruiwen Zhou, Jiaqian Wang, Shengtao Zhang, Wei Zhang, Ying Wen, Zhiyu Li, Feiyu Xiong, Bo Tang, Weinan Zhang, Muning Wen 5/15/2026

MemQ: Integrating Q-Learning into Self-Evolving Memory Agents over Provenance DAGs

MemQ framework integrating Q-learning into LLM agent memory systems using TD eligibility traces to propagate credit backward through memory provenance DAGs.

Ax Ying Chen, Lihuang Fang, Rui Jiang, Mingxu Wang, Zhifeng Gu, Lei Yi, Jie Chen 5/15/2026

Done, But Not Sure: Disentangling World Completion from Self-Termination in Embodied Agents

VIGIL evaluation framework for embodied agents that independently measures terminal commitment, distinguishing task completion failure modes previously conflated in benchmarks.

Ax Jinchao Hu, Meizhi Zhong, Kehai Chen, Min Zhang 5/15/2026

SearchSkill: Teaching LLMs to Use Search Tools with Evolving Skill Banks

Framework teaching LLMs to use search tools effectively by making query planning explicit through reusable search skills, improving open-domain question answering performance.

Ax Xiao Jia 5/15/2026

Unpredictability dissociates from structured control in language agents

Research on control mechanisms in language agents, testing whether stochastic sampling can substitute for structured control components coupling memory, reasoning, and action selection.

Ax Jianghan Shen, Siqi Luo, Xinyu Cheng, Jing Xiong, Yue Li, Jiyao Liu, Jiashi Lin, Yirong Chen, Junjun He 5/15/2026

CuSearch: Curriculum Rollout Sampling via Search Depth for Agentic RAG

CuSearch proposes curriculum sampling for training agentic RAG systems via reinforcement learning, optimizing policies based on trajectory search depth rather than uniform sampling.

Ax Ali Al-Lawati, Nafis Tripto, Abolfazl Ansari, Jason Lucas, Suhang Wang, Dongwon Lee 5/15/2026

Moltbook Moderation: Uncovering Hidden Intent Through Multi-Turn Dialogue

BOT-MOD moderation system detecting malicious agent intent in multi-agent systems through multi-turn dialogue analysis beyond content filtering.

Ax Yucheng Guo, Yongjian Guo, Zhong Guan, Wen Huang, Haoran Sun, Haodong Yue, Xiaolong Xiang, Shuai Di, Zhen Sun, Luqiao Wang, Junwu Xiong, Yicheng Gong 5/15/2026

D-VLA: A High-Concurrency Distributed Asynchronous Reinforcement Learning Framework for Vision-Language-Action Models

D-VLA distributed asynchronous RL framework for training Vision-Language-Action models at scale, addressing resource conflicts in embodied AI.

Ax Kangning Zhang, Shuai Shao, Qingyao Li, Jianghao Lin, Lingyue Fu, Shijian Wang, Wenxiang Jiao, Yuan Lu, Weiwen Liu, Weinan Zhang, Yong Yu 5/15/2026

MMSkills: Towards Multimodal Skills for General Visual Agents

MMSkills framework for reusable multimodal skills in visual agents, encoding procedural knowledge across visual, textual and executable modalities.

Ax Hongxuan Wang, Xiaocong Li, Lihao Zheng, Adrish Bhaumik, Prahlad Vadakkepat 5/15/2026

Safe Bayesian Optimization for Complex Control Systems via Additive Gaussian Processes

Safe Bayesian optimization for controller tuning in robotics via additive Gaussian processes with safety guarantees for physical systems.

Ax Fangchen Song, Ashish Agarwal, Wen Wen 5/15/2026

The Impact of Generative AI on Collaborative Open-Source Software Development: Evidence from GitHub Copilot

Empirical study on GitHub Copilot's impact on collaborative open-source software development productivity and participation patterns.

Ax Zhiliang Chen, Gregory Kang Ruey Lau, Chuan-Sheng Foo, Bryan Kian Hsiang Low 5/15/2026

DUET: Optimizing Training Data Mixtures via Feedback from Unseen Evaluation Tasks

DUET optimizes LLM training data mixtures using feedback from unseen evaluation tasks without access to task-specific data.

Ax Jathurshan Pradeepkumar, Xihao Piao, Zheng Chen, Jimeng Sun 5/15/2026

Tokenizing Single-Channel EEG with Time-Frequency Motif Learning

TFM-Tokenizer learns time-frequency motif vocabulary from single-channel EEG signals for tokenization in foundation models using dual-path architecture.

Ax Tianneng Shi, Jingxuan He, Zhun Wang, Hongwei Li, Linyu Wu, Wenbo Guo, Dawn Song 5/15/2026

Progent: Securing AI Agents with Privilege Control

Progent framework securing AI agents with privilege control against prompt injection attacks through dynamic security requirements and tool access management.