Ax Alexandre Cristov\~ao Maiorano 3/31/2026

LLM Readiness Harness: Evaluation, Observability, and CI Gates for LLM/RAG Applications

Readiness harness for LLM/RAG applications combining evaluation, OpenTelemetry observability, and CI gates with Pareto frontier readiness scores.

Ax Jyotsana Khatri, Manasi Patwardhan 3/31/2026

Defend: Automated Rebuttals for Peer Review with Minimal Author Guidance

Defend: Automated peer review rebuttal generation using LLMs with structured reasoning for targeted refutation and factual grounding.

Ax Jakub Mas{\l}owski, Jaros{\l}aw A. Chudziak 3/31/2026

Heterogeneous Debate Engine: Identity-Grounded Cognitive Architecture for Resilient LLM-Based Ethical Tutoring

Identity-grounded multi-agent LLM system with debate engine for resilient ethical tutoring addressing semantic drift and logical deterioration.

Ax Yitao Li 3/31/2026

Greedy Is a Strong Default: Agents as Iterative Optimizers

LLM agents replace random proposal generators in classical optimization algorithms; evaluated on discrete, mixed, and continuous optimization tasks.

Ax Mahesh Natarajan, Xiaoye Li, Weiqun Zhang 3/31/2026

AstraAI: LLMs, Retrieval, and AST-Guided Assistance for HPC Codebases

AstraAI: CLI framework integrating LLMs with RAG and AST-based analysis for context-aware code generation in HPC scientific codebases.

Ax Jacky Liang 3/31/2026

The Novelty Bottleneck: A Framework for Understanding Human Effort Scaling in AI-Assisted Work

Novelty Bottleneck model formalizing human-AI collaboration limits: irreducible serial component from tasks requiring human judgment follows Amdahl's Law.

Ax Wei Wang, Tianyu Shi, Shuai Zhang, Boyang Xia, Zequn Xie, Chenyu Zeng, Qi Zhang, Lynn Ai, Yaqi Yu, Kaiming Zhang, Feiyue Tang 3/31/2026

PeopleSearchBench: A Multi-Dimensional Benchmark for Evaluating AI-Powered People Search Platforms

PeopleSearchBench: Open-source benchmark evaluating AI-powered people search platforms across recruiting, sales prospecting, and expert discovery use cases.

Ax Jean Douglas Carvalho, Hugo Taciro Kenji, Ahmad Mohammad Saber, Glaucia Melo, Max Mauro Dias Santos, Deepa Kundur 3/31/2026

Dual-Stage LLM Framework for Scenario-Centric Semantic Interpretation in Driving Assistance

Dual-stage LLM framework for scenario-centric semantic risk interpretation in autonomous driving assistance systems with reproducible auditing.

Ax Jin Huang, Jie Yang, XinLei Zhou, Qihao Liu, Liang Gao, Xinyu Li 3/31/2026

DSevolve: Enabling Real-Time Adaptive Scheduling on Dynamic Shop Floor with LLM-Evolved Heuristic Portfolios

DSevolve: Industrial scheduling framework using LLM to evolve diverse heuristic portfolios for adaptive rule selection in dynamic manufacturing environments.

Ax Kaikai Zhang, Xiang Wang, Haoluo Zhao, Nan Chen, Mengyang Yu Jing-Jia Luo, Tao Song, Fan Meng 3/31/2026

TianJi:An autonomous AI meteorologist for discovering physical mechanisms in atmospheric science

TianJi: Autonomous AI system using LLMs to discover physical causal mechanisms in atmospheric science beyond statistical weather forecasting.

Ax Adam Haile 3/31/2026

SkyNet: Belief-Aware Planning for Partially-Observable Stochastic Games

SkyNet: Extends MuZero model-based RL to partially observable stochastic multi-player games with belief-aware planning under hidden state uncertainty.

Ax Yin Cheng, Liao Zhou, Xiyu Liang, Dihao Luo, Tewei Lee, Kailun Zheng, Weiwei Zhang, Mingchen Cai, Jian Dong, Andy Zhang 3/31/2026

Let the Agent Steer: Closed-Loop Ranking Optimization via Influence Exchange

Sortify: Agent-based closed-loop ranking optimization for recommendation systems treating ranking as influence allocation problem with online-offline metric calibration.

Ax Swarna Kamal Paul, Shubhendu Sharma, Nitin Sareen 3/31/2026

GAAMA: Graph Augmented Associative Memory for Agents

GAAMA: Graph-based long-term memory system for AI agents maintaining personalized, multi-session behavior through associative memory structure instead of flat retrieval.

Ax Camilo Chac\'on Sartori, Jos\'e H. Garc\'ia, Andrei Voicu Tomut, Christian Blum 3/31/2026

GEAKG: Generative Executable Algorithm Knowledge Graphs

GEAKG: Knowledge graphs framework for representing procedural algorithm knowledge as executable, learnable structures for domain-agnostic problem solving.

Ax Yongkang Du, Xiaohan Zou, Minhao Cheng, Lu Lin 3/31/2026

CARV: A Diagnostic Benchmark for Compositional Analogical Reasoning in Multimodal LLMs

CARV: Benchmark dataset with 5,500 samples evaluating compositional analogical reasoning capabilities in multimodal LLMs, testing higher-order intelligence composition.

Ax Yifan Wang, Bolian Li, David Cho, Ruqi Zhang, Fanping Sui, Ananth Grama 3/31/2026

SARL: Label-Free Reinforcement Learning by Rewarding Reasoning Topology

SARL: Label-free reinforcement learning method for improving reasoning in LLMs without requiring verifiable rewards, addressing open-ended domains with ambiguous correctness.

Ax Xujia Li, Xin Li, Junquan Huang, Beirong Cui, Zibin Wu, Lei Chen 3/31/2026

HeteroHub: An Applicable Data Management Framework for Heterogeneous Multi-Embodied Agent System

Framework for managing heterogeneous data in multi-embodied agent systems with diverse capabilities operating in dynamic environments.

Ax Edward Wijaya 3/31/2026

What an Autonomous Agent Discovers About Molecular Transformer Design: Does It Transfer?

AI agent uses architecture search across 3,106 experiments to test whether molecular sequences need different transformer designs than language models.

Ax Taeyun Roh, Eun-yeong Jo, Wonjune Jang, Jaewoo Kang 3/31/2026

When Choices Become Priors: Contrastive Decoding for Scientific Figure Multiple-Choice QA

Study of bias in multimodal LLMs on scientific figure QA tasks where answer choices influence model predictions despite contradictory visual evidence.

Ax Rohan Pandey, Eric Ye, Michael Li 3/31/2026

Beyond the Answer: Decoding the Behavior of LLMs as Scientific Reasoners

Research on how LLMs perform scientific reasoning, examining internal heuristics and the effect of prompting on reasoning processes.

Ax Yoonho Lee, Roshen Nair, Qizheng Zhang, Kangwook Lee, Omar Khattab, Chelsea Finn 3/31/2026

Meta-Harness: End-to-End Optimization of Model Harnesses

arXiv paper: Meta-Harness, automated optimization system searching over LLM harness code to improve system performance beyond model weights.

Ax Yuang Wei, Ruijia Li, Bo Jiang 3/31/2026

SLOW: Strategic Logical-inference Open Workspace for Cognitive Adaptation in AI Tutoring

arXiv paper: SLOW, LLM-based tutoring system with dedicated reasoning workspace for cognitive diagnosis and pedagogical decision-making.

Ax Jiacheng Wang, Jinbin Huang 3/31/2026

Reward Hacking as Equilibrium under Finite Evaluation

arXiv paper proving reward hacking is structural equilibrium in optimized AI agents under finite evaluation, regardless of alignment method.

Ax Siyuan Ma, Bo Gao, Zikai Xiao, Hailong Wang, Xinlei Yu, Rui Qian, Jiayu Qian, Luqi Gong, Yang Liu 3/31/2026

CoT2-Meta: Budgeted Metacognitive Control for Test-Time Reasoning

arXiv paper: CoT2-Meta, training-free metacognitive reasoning framework combining chain-of-thought with meta-level control over reasoning trajectories.

Ax Zehua Han, Jing Xiao, Yiqi Duan, Mengyu Xiang, Yuheng Ji, Xiaolong Zheng, Chenghanyu Zhang, Zhendong She, Junyu Shen, Dingwei Tan, Shichu Sun, Zhou Cong, Mingxuan Liu, Fengxiang Wang, Jinping Sun, Yangang Sun 3/31/2026

PReD: An LLM-based Foundation Multimodal Model for Electromagnetic Perception, Recognition, and Decision

arXiv paper: PReD, foundation multimodal model for electromagnetic domain covering perception, recognition, and decision-making.

Ax Yujie Zhang, Weikang Yuan, Zhuoren Jiang, Pengwei Yan 3/31/2026

EpiPersona: Persona Projection and Episode Coupling for Pluralistic Preference Modeling

arXiv paper: EpiPersona framework for pluralistic LLM alignment, explicitly coupling stable personas with episode-specific preference factors.

Ax David K. Johansson 3/31/2026

Reasoning as Energy Minimization over Structured Latent Trajectories

arXiv paper: Energy-Based Reasoning via Structured Latent Planning (EBRM), models reasoning as gradient-based optimization over multi-step latent trajectories.

Ax Thomas Van Mullem, Bart Mesuere, Peter Dawyndt 3/31/2026

Evaluating LLMs for Answering Student Questions in Introductory Programming Courses

arXiv study evaluating LLMs for answering student programming questions, addressing balance between pedagogical hints and complete solutions.

Ax Julio C. Serrano. Joonas Kevari, Rumy Narayan 3/31/2026

A Multi-Agent Rhizomatic Pipeline for Non-Linear Literature Analysis

arXiv paper: Rhizomatic Research Agent (V3), multi-agent pipeline for non-linear literature analysis in social sciences using Deleuzian ontology.

Ax Kangkang Sun, Jun Wu, Jianhua Li, Minyi Guo, Xiuzhen Che, Jianwei Huang 3/31/2026

CoE: Collaborative Entropy for Uncertainty Quantification in Agentic Multi-LLM Systems

arXiv paper proposing Collaborative Entropy (CoE), information-theoretic metric for semantic uncertainty quantification in multi-LLM agentic systems.

Ax Yipeng Yu 3/31/2026

Deep Research of Deep Research: From Transformer to Agent, From AI to AI for Science

Survey tracing evolution of LLMs from transformers to agents, analyzing deep research as a prototype vertical application demonstrating progression from QA to agentic tool use.

Ax Alkis Sygkounas, Rishi Hazra, Andreas Persson, Pedro Zuidberg Dos Martires, Amy Loutfi 3/31/2026

COvolve: Adversarial Co-Evolution of Large-Language-Model-Generated Policies and Environments via Two-Player Zero-Sum Game

COvolve co-evolves LLM-generated agent policies and environments as executable Python code using adversarial game dynamics to enable continual learning and improved generalization.

Ax Doan Nam Long Vu, Simone Balloccu 3/31/2026

The Scaffold Effect: How Prompt Framing Drives Apparent Multimodal Gains in Clinical VLM Evaluation

Study evaluates 12 open-weight vision-language models on clinical neuroimaging tasks, revealing that performance gains may reflect prompt artifacts rather than genuine multimodal evidence integration.

Ax Fangda Ye, Yuxin Hu, Pengxiang Zhu, Yibo Li, Ziqi Jin, Yao Xiao, Yibo Wang, Lei Wang, Zhen Zhang, Lu Wang, Yue Deng, Bin Wang, Yifan Zhang, Liangcai Su, Xinyu Wang, He Zhao, Chen Wei, Qiang Ren, Bryan Hooi, An Bo, Shuicheng Yan, Lidong Bing 3/31/2026

MiroEval: Benchmarking Multimodal Deep Research Agents in Process and Outcome

MiroEval benchmarks multimodal deep research agents by evaluating both research process and outcomes, addressing limitations of existing benchmarks with real-world query complexity and multimodal coverage.

Ax Davide Di Gioia 3/31/2026

Entropic Claim Resolution: Uncertainty-Driven Evidence Selection for RAG

Entropic Claim Resolution introduces a novel RAG inference method that uses uncertainty-driven evidence selection to handle conflicting information and query ambiguity beyond relevance-based retrieval.

Ax Adam Laabs 3/31/2026

T-Norm Operators for EU AI Act Compliance Classification: An Empirical Comparison of Lukasiewicz, Product, and G\"odel Semantics in a Neuro-Symbolic Reasoning System

Pilot study comparing t-norm operators in neuro-symbolic reasoning system for EU AI Act compliance classification of AI systems.

Ax Hongtao Wu, Boyun Zheng, Dingjie Song, Yu Jiang, Jianfeng Gao, Lei Xing, Lichao Sun, Yixuan Yuan 3/31/2026

Towards a Medical AI Scientist

Medical AI Scientist: autonomous system generating medical hypotheses, conducting experiments, and drafting manuscripts grounded in clinical evidence.

Ax Han Wang, Yifan Sun, Brian Ko, Mann Talati, Jiawen Gong, Zimeng Li, Naicheng Yu, Xucheng Yu, Wei Shen, Vedant Jolly, Huan Zhang 3/31/2026

MonitorBench: A Comprehensive Benchmark for Chain-of-Thought Monitorability in Large Language Models

MonitorBench: comprehensive open-source benchmark studying chain-of-thought monitorability and causal responsibility in LLM reasoning.

Ax Ziqi Miao, Haonan Jia, Lijun Li, Chen Qian, Yuan Xiong, Wenting Yan, Jing Shao 3/31/2026

Seeing with You: Perception-Reasoning Coevolution for Multimodal Reasoning

Perception-Reasoning Coevolution framework decoupling perception and reasoning optimization in multimodal LLM reasoning with verifiable rewards.

Ax Lara Russell-Lasalandra, Hudson Golino, Luis Eduardo Garrido, Alexander P. Christensen 3/31/2026

The Ultimate Tutorial for AI-driven Scale Development in Generative Psychometrics: Releasing AIGENIE from its Bottle

AIGENIE R package automating psychological scale development using LLM text generation with network psychometric methods.

Ax Rongjin Li, Zichen Tang, Xianghe Wang, Xinyi Hu, Zhengyu Wang, Zhengyu Lu, Yiling Huang, Jiayuan Chen, Weisheng Tan, Jiacheng Liu, Zhongjun Yang, Haihong E 3/31/2026

Not Search, But Scan: Benchmarking MLLMs on Scan-Oriented Academic Paper Reasoning

ScanPaper benchmark evaluating MLLMs on scan-oriented academic paper reasoning beyond search-oriented retrieval tasks.

Ax Songjun Tu, Chengdong Xu, Qichao Zhang, Yaocheng Zhang, Xiangyuan Lan, Linjing Li, Dongbin Zhao 3/31/2026

Dynamic Dual-Granularity Skill Bank for Agentic RL

D2Skill: dynamic dual-granularity skill bank organizing reusable experience into task and step skills for reinforcement learning agents.

Ax Hadi Mohammadi, Ayoub Bagheri 3/31/2026

Exploring Cultural Variations in Moral Judgments with Large Language Models

Study comparing whether smaller and larger LLMs capture culturally diverse moral values from World Values Survey and Pew Research data.

Ax Yadi Cao, Sicheng Lai, Jiahe Huang, Yang Zhang, Zach Lawrence, Rohan Bhakta, Izzy F. Thomas, Mingyun Cao, Chung-Hao Tsai, Zihao Zhou, Yidong Zhao, Hao Liu, Alessandro Marinoni, Alexey Arefiev, Rose Yu 3/31/2026

SimulCost: A Cost-Aware Benchmark and Toolkit for Automating Physics Simulations with LLMs

SimulCost: cost-aware benchmark for LLM agents in physics simulations accounting for tool-use costs like simulation time and experimental resources.

Ax Sun Xu, Tongkai Xu, Baiheng Xie, Li Huang, Qiang Gao, Kunpeng Zhang 3/31/2026

M-RAG: Making RAG Faster, Stronger, and More Efficient

M-RAG: improved RAG system addressing retrieval noise and inefficiency through better chunking strategies for long-context LLMs.

Ax Zihang Li, Wenjun Liu, Yikun Zong, Jiawen Tao, Siying Dai, Songcheng Ren, Zirui Liu, Yanbing Jiang, Tong Yang 3/31/2026

Bridge-RAG: An Abstract Bridge Tree Based Retrieval Augmented Generation Algorithm With Cuckoo Filter

Bridge-RAG: novel RAG framework using abstract bridge trees and cuckoo filters to improve retrieval accuracy and efficiency for LLM generation.

Ax Yuan Hu, ZhiYu Cao, PeiFeng Li, QiaoMing Zhu 3/31/2026

ReCQR: Incorporating conversational query rewriting to improve Multimodal Image Retrieval

ReCQR: conversational query rewriting approach for multimodal image retrieval handling long text and unclear user expressions.

Ax Talita de Paula Cypriano de Souza, Shruti Mehta, Matheus Arataque Uema, Luciano Bernardes de Paula, Seiji Isotani 3/31/2026

Can AI be a Teaching Partner? Evaluating ChatGPT, Gemini, and DeepSeek across Three Teaching Strategies

Comparative empirical study evaluating ChatGPT, Gemini, and DeepSeek as teaching agents across pedagogical strategies.

Ax Michelle Vaccaro, Jaeyoon Song, Abdullah Almaatouq, Michiel A. Bakker 3/31/2026

Evaluating Human-AI Safety: A Framework for Measuring Harmful Capability Uplift

Framework proposing harmful capability uplift as core safety metric for measuring marginal increases in user ability to cause harm with frontier AI models.

Ax J\'er\'emy Barghorn, Anna Sotnikova, Sacha Friedli, Antoine Bosselut 3/31/2026

AI Meets Mathematics Education: A Case Study on Supporting an Instructor in a Large Mathematics Class with Context-Aware AI

Case study of context-aware AI system supporting calculus instruction by answering student questions on discussion forums.