Ax Wenyuan Zhang, Shuaiyi Nie, Xinghua Zhang, Zefeng Zhang, Tingwen Liu 5/5/2026

Exploring the System 1 Thinking Capability of Large Reasoning Models

Analysis of system 1 thinking capability in large reasoning models, evaluating intuitive responses with minimal tokens for efficiency and difficulty awareness.

Ax Luca Castri, Gloria Beraldo, Nicola Bellotto 5/5/2026

Causality-enhanced Decision-Making for Autonomous Mobile Robots in Dynamic Environments

Research on causal inference for autonomous mobile robots in dynamic environments, modeling human behavior and interactions.

Ax Ziyuan Zhang, Darcy Wang, Ningyuan Chen, Rodrigo Mansur, Vahid Sarhangian 5/5/2026

Comparing Exploration-Exploitation Strategies of LLMs and Humans: Insights from Standard Multi-armed Bandit Experiments

Study comparing exploration-exploitation tradeoffs in LLMs versus humans using multi-armed bandit experiments for sequential decision-making.

Ax Jatin Gupta, Akhil Sharma, Saransh Singhania, Ali Imam Abidi 5/5/2026

Lightweight Domain Adaptation of a Large Language Model for Legal Assistance in the Indian Context

Legal Assist AI: Lightweight domain-adapted framework for providing legal assistance in Indian context using efficient LLM fine-tuning.

Ax Zexi Liu, Jingyi Chai, Xinyu Zhu, Shuo Tang, Rui Ye, Bo Zhang, Lei Bai, Siheng Chen 5/5/2026

ML-Agent: Reinforcing LLM Agents for Autonomous Machine Learning Engineering

ML-Agent: Framework reinforcing LLM agents for autonomous machine learning engineering, moving beyond prompt-based paradigms for better generalization.

Ax Kundan Krishna, Joseph Y Cheng, Charles Maalouf, Leon A Gatys 5/5/2026

Disentangled Safety Adapters Enable Efficient Guardrails and Flexible Inference-Time Alignment

Disentangled Safety Adapters: Framework decoupling safety computations from base models using lightweight adapters for efficient guardrails and flexible alignment.

Ax Jiacong Wang, Zijian Kang, Haochen Wang, Haiyong Jiang, Jiawen Li, Bohong Wu, Ya Wang, Jiao Ran, Xiao Liang, Chao Feng, Jun Xiao 5/5/2026

VGR: Visual Grounded Reasoning

VGR: Framework for visual grounded chain-of-thought reasoning that performs reasoning in visual space rather than pure language, for complex visual tasks.

Ax Lehan He, Zeren Chen, Zhe Zhang, Xiang Gao, Lu Sheng 5/5/2026

Effective LLM Code Refinement via Property-Oriented and Structurally Minimal Feedback

Method for refining LLM-generated code through property-oriented, minimal feedback instead of test quantity, improving functional correctness.

Ax Rahul Ramachandran, Ali Garjani, Roman Bachmann, Andrei Atanov, O\u{g}uzhan Fatih Kar, Amir Zamir 5/5/2026

How Well Does GPT-4o Understand Vision? Evaluating Multimodal Foundation Models on Standard Computer Vision Tasks

Benchmark comparing multimodal foundation models (GPT-4o, Gemini, Claude, Qwen, Llama) on standard computer vision tasks beyond question answering.

Ax Yiran Wu, Mauricio Velazco, Andrew Zhao, Manuel Ra\'ul Mel\'endez Luj\'an, Srisuma Movva, Yogesh K Roy, Quang Nguyen, Roberto Rodriguez, Qingyun Wu, Michael Albada, Julia Kiseleva, Anand Mudgerikar 5/5/2026

ExCyTIn-Bench: Evaluating LLM agents on Cyber Threat Investigation

ExCyTIn-Bench: First benchmark for evaluating LLM agents on autonomous cyber threat investigation using multi-hop evidence chains from security logs.

Ax Zhaomin Wu, Mingzhe Du, See-Kiong Ng, Bingsheng He 5/5/2026

Beyond Prompt-Induced Lies: Investigating LLM Deception on Benign Prompts

Study investigating intentional deception in LLMs on benign prompts without explicit hidden objectives, assessing trustworthiness for decision-making tasks.

Ax Anirudh Iyengar Kaniyar Narayana Iyengar, Srija Mukhopadhyay, Adnan Qidwai, Shubhankar Singh, Dan Roth, Vivek Gupta 5/5/2026

InterChart: Benchmarking Visual Reasoning Across Decomposed and Distributed Chart Information

InterChart benchmark evaluating vision-language models on reasoning across multiple related charts for scientific, financial, and policy analysis tasks.

Ax Arshia Akhavan, Alireza Hosseinpour, Abbas Heydarnoori, Hamid Bagheri, Mehdi Keshani 5/5/2026

LinkAnchor: An Autonomous LLM-Based Agent for Issue-to-Commit Link Recovery

LinkAnchor: Autonomous LLM agent for recovering issue-to-commit links in software repositories, addressing low linking rates on GitHub.

Ax Diane Tchuindjo, Omar Khattab 5/5/2026

Reasoning-Intensive Regression

Framework for using LLMs to perform reasoning-intensive regression tasks like rubric-based scoring and dense reward modeling, beyond standard text analysis.

Ax Jie Li, Youyang Hou, Laura Lin, Ruihao Zhu, Hancheng Cao, Abdallah El Ali 5/5/2026

Vibe Coding in Product Teams: Reconfiguring AI-Assisted Workflows, Prototyping, and Collaboration

Qualitative study of how generative AI reshapes product development workflows through natural language-to-code translation, based on 22 interviews across teams.

Ax Zhaomin Wu, Haodong Zhao, Ziyang Wang, Jizhou Guo, Qian Wang, Bingsheng He 5/5/2026

LLM DNA: Tracing Model Evolution via Functional Representations

Method using functional representations to trace evolutionary relationships between millions of LLMs, addressing undocumented fine-tuning and adaptation chains.

Ax Kai-Wei Chang, En-Pei Hu, Chun-Yi Kuan, Wenze Ren, Wei-Chih Chen, Guan-Ting Lin, Yu Tsao, Shao-Hua Sun, Hung-yi Lee, James Glass 5/5/2026

Game-Time: Evaluating Temporal Dynamics in Spoken Language Models

Benchmark framework (Game-Time) for evaluating temporal dynamics capabilities in conversational spoken language models, including timing and simultaneous speech handling.

Ax Chenhui Xu, Fuxun Yu, Michael J. Bianco, Jacob Kovarskiy, Raphael Tang, Qi Zhang, Zirui Xu, Will LeVine, Brandon Dubbs, Heming Liao, Cassandra Burgess, Suvam Bag, Jay Patravali, Rupanjali Kukal, Mikael Figueroa, Rishi Madhok, Nikolaos Karianakis, Jinjun Xiong 5/5/2026

Unlocking Zero-Shot Geospatial Reasoning via Indirect Rewards

Research on training vision-language models for geospatial reasoning using indirect rewards from metadata to overcome supervision scarcity in rare domains.

Ax Lorenzo Nikiforos, Luciano Prono, Charalampos Antoniadis, Fabio Pareschi, Riccardo Rovatti, Gianluca Setti 5/5/2026

Vanishing Contributions: A Unified Framework for Smooth and Iterative Model Compression

Unified framework for iterative neural network compression combining pruning, quantization, and low-rank decomposition techniques with gradual accuracy preservation.

Ax Charidimos Papadakis, Angeliki Dimitriou, Giorgos Filandrianos, Maria Lymperaiou, Konstantinos Thomas, Giorgos Stamou 5/5/2026

ATLAS: Adaptive Trading with LLM AgentS Through Dynamic Prompt Optimization and Multi-Agent Coordination

ATLAS: framework for deploying LLM agents in autonomous trading with dynamic prompt optimization, multi-agent coordination, and late-arriving noisy rewards.

Ax Qingyao Ai, Yichen Tang, Changyue Wang, Jianming Long, Weihang Su, Yiqun Liu 5/5/2026

MemoryBench: A Benchmark for Memory and Continual Learning in LLM Systems

MemoryBench: benchmark for evaluating memory and continual learning capabilities in LLM systems, addressing limitations of scaling-only approaches.

Ax Jackson Hassell, Dan Zhang, Hannah Kim, Tom Mitchell, Estevam Hruschka 5/5/2026

Learning from Supervision with Semantic and Episodic Memory: A Reflective Approach to Agent Adaptation

Memory-augmented framework enabling LLM agents to learn classification functions from labeled examples without parameter updates using episodic memory and critiques.

Ax Md. Mehedi Hasan, Sk Tanzir Mehedi, Ziaur Rahman, Rafid Mostafiz, Md. Abir Hossain 5/5/2026

Sentra-Guard: A Real-Time Multilingual Defense Against Adversarial LLM Prompts

Sentra-Guard: real-time defense system detecting jailbreak and prompt injection attacks on LLMs using FAISS embeddings and fine-tuned classifiers.

Ax Feijie Wu, Weiwu Zhu, Yuxiang Zhang, Soumya Chatterjee, Jiarong Zhu, Fan Mo, Rong Luo, Jing Gao 5/5/2026

PORTool: Importance-Aware Policy Optimization with Rewarded Tree for Multi-Tool-Integrated Reasoning

PORTool: algorithm for training multi-tool LLM agents using importance-weighted policy optimization to resolve credit-assignment ambiguity in tool-use decisions.

Ax Hatim Chergui, Farhad Rezazadeh, Mehdi Bennis, Merouane Debbah, Christos Verikoukis 5/5/2026

LLM-Based Agentic Negotiation for 6G: Addressing Uncertainty Neglect and Tail-Event Risk

Risk-aware LLM-based agentic framework for 6G network negotiation addressing uncertainty neglect and tail-event risks.

Ax Th\'eo Cavignac (Research Center Future Energy Materials and Systems of the University Alliance Ruhr and ICAMS, Ruhr University Bochum, Bochum, Germany), Jonathan Schmidt (Department of Materials, ETH Z\"urich, Z\"urich, Switzerland), Pierre-Paul De Breuck (Research Center Future Energy Materials and Systems of the University Alliance Ruhr and ICAMS, Ruhr University Bochum, Bochum, Germany), Antoine Loew (Research Center Future Energy Materials and Systems of the University Alliance Ruhr and ICAMS, Ruhr University Bochum, Bochum, Germany), Tiago F. T. Cerqueira (CFisUC, Department of Physics, University of Coimbra, Coimbra, Portugal), Hai-Chen Wang (Research Center Future Energy Materials and Systems of the University Alliance Ruhr and ICAMS, Ruhr University Bochum, Bochum, Germany), Anton Bochkarev (ICAMS, Ruhr-Universit\"at Bochum and ACEworks GmbH, Bochum, Germany), Yury Lysogorskiy (ICAMS, Ruhr-Universit\"at Bochum and ACEworks GmbH, Bochum, Germany), Aldo H. Romero (Department of Physics, West Virginia University, Morgantown, USA), Ralf Drautz (ICAMS, Ruhr-Universit\"at Bochum and ACEworks GmbH, Bochum, Germany), Silvana Botti (Research Center Future Energy Materials and Systems of the University Alliance Ruhr and ICAMS, Ruhr University Bochum, Bochum, Germany), Miguel A. L. Marques (Research Center Future Energy Materials and Systems of the University Alliance Ruhr and ICAMS, Ruhr University Bochum, Bochum, Germany) 5/5/2026

AI-Driven Expansion and Application of the Alexandria Database

Multi-stage computational materials discovery workflow combining generative models and graph neural networks generating 119M candidate structures.

Ax Jordan Taylor, William Agnew, Maarten Sap, Sarah E. Fox, Haiyi Zhu 5/5/2026

The Algorithmic Gaze of Image Quality Assessment: An Audit and Trace Ethnography of the LAION-Aesthetics Predictor

Audit of LAION-Aesthetics Predictor examining whose cultural values and aesthetic preferences are encoded in image quality models.

Ax Nigam H. Shah, Nerissa Ambers, Abby Pandya, Timothy Keyes, Juan M. Banda, Srikar Nallan, Carlene Lugtu, Artem A. Trotsyuk, Suhana Bedi, Alyssa Unell, Miguel Fuentes, Francois Grolleau, Sneha S. Jain, Jonathan Chen, Devdutta Dash, Danton Char, Aditya Sharma, Duncan McElfresh, Patrick Scully, Vishanthan Kumar, Clancy Dennis, Connor OBrien, Satchi Mouniswamy, Elvis Jones, Krishna Jasti, Gunavathi Mannika Lakshmanan, Sree Ram Akula, Varun Kumar Singh, Ramesh Rajmanickam, Sudhir Sinha, Vicky Zhou, Xu Wang, Bilal Mawji, Joshua Ge, Wencheng Li, Travis Lyons, Jarrod Helzer, Vikas Kakkar, Ramesh Powar, Darren Batara, Cheryl Cordova, William Frederick III, Olivia Tang, Phoebe Morgan, April S. Liang, Stephen P. Ma, Shivam Vedak, Dong-han Yao, Akshay Swaminathan, Mehr Kashyap, Brian Ng, Jamie Hellman, Nikesh Kotecha, Christopher Sharp, Gretchen Brown, Christian Lindmark, Anurang Revri, Michael A. Pfeffer 5/5/2026

Adoption and Use of LLMs at an Academic Medical Center

ChatEHR system enabling LLM integration with electronic health records for clinical documentation automations and interactive use.

Ax Lakshan Cooray, Deshan Sumanathilaka, Pattigadapa Venkatesh Raju 5/5/2026

Can Small Language Models Handle Context-Summarized Multi-Turn Customer-Service QA? A Synthetic Data-Driven Comparative Evaluation

Comparative evaluation of small language models on multi-turn customer-service QA using synthetic data in resource-constrained settings.

Ax Michael A. Lepori, Tal Linzen, Ann Yuan, Katja Filippova 5/5/2026

Language Models Struggle to Use Representations Learned In-Context

Research showing LLMs struggle to utilize representations learned in-context, limiting few-shot adaptation capabilities.

Ax Abdullah Arafat Miah, Kevin Vu, Yu Bi 5/5/2026

BadSNN: Backdoor Attacks on Spiking Neural Networks via Adversarial Spiking Neuron

Backdoor attack method targeting spiking neural networks via adversarial manipulation of spiking neuron hyperparameters.

Ax Aydin Ayanzadeh, Prakhar Dixit, Sadia Kamal, Milton Halem 5/5/2026

WildfireVLM: AI-powered Analysis for Early Wildfire Detection and Risk Assessment Using Satellite Imagery

Vision-language model framework for early wildfire detection and risk assessment combining satellite imagery analysis with AI.

Ax Edward Izgorodin 5/5/2026

Semantic Level of Detail for Knowledge Graphs: Discovering Abstraction Boundaries via Spectral Heat Diffusion

Method using spectral heat diffusion to identify abstraction boundaries in knowledge graphs for continuous resolution control.

Ax Chenglei Shen, Teng Shi, Weijie Yu, Xiao Zhang, Jun Xu 5/5/2026

GenRecEdit: Adapting Model Editing for Generative Recommendation with Cold-Start Items

Model editing technique for generative recommendation systems addressing cold-start item collapse without full retraining.

Ax Gregory N. Frank 5/5/2026

Detection Is Cheap, Routing Is Learned: Why Refusal-Based Alignment Evaluation Fails

Study of alignment evaluation in LLMs examining concept detection vs. behavioral routing using Chinese-origin models and ablation techniques.

Ax Gregory N. Frank 5/5/2026

How Alignment Routes: Localizing, Scaling, and Controlling Policy Circuits in Language Models

Research localizing policy routing mechanisms in alignment-trained LLMs, identifying gate and amplifier heads controlling refusal behavior.

Ax Andoni Irazusta Garmendia 5/5/2026

A First Guess is Rarely the Final Answer: Learning to Search in the Traveling Salesperson Problem

Neural improvement methods for TSP that learn search policies applying local modifications to candidate solutions rather than single-shot outputs.

Ax Tao Li, Kaiyuan Hou, Tuan Vinh, Monika Raj, Zhichun Guo, Carl Yang 5/5/2026

Reinforcement Learning with LLM-Guided Action Spaces for Synthesizable Lead Optimization

Framework combining reinforcement learning with LLM-guided action spaces for optimizing drug molecules while ensuring synthesizability.

Ax Wei Chen, Yubing Wu, Junmei Yang, Delu Zeng, Qibin Zhao, John Paisley, Min Chen, Zhou Wang 5/5/2026

Towards Disentangled Preference Optimization Dynamics: Suppress the Loser, Preserve the Winner

Unified framework for preference optimization revealing shared decomposition across margin-based methods to preserve chosen responses while suppressing rejected ones.

Ax Emil Ryd, Henning Bartsch, Julian Stastny, Joe Benton, Vivek Hebbar 5/5/2026

Removing Sandbagging in LLMs by Training with Weak Supervision

Technique using weak supervision to prevent sandbagging in LLMs, eliciting best performance without reliable verification of output quality.

Ax Xiaohang Yu, Hejia Geng, Xinmeng Zeng, William Knottenbelt 5/5/2026

SUDP: Secret-Use Delegation Protocol for Agentic Systems

SUDP: delegation protocol preventing credential exposure and account compromise in agentic systems accessing user secrets and APIs.

Ax Wenshuo Zhao (May), Qi Zhu (May), Xingshan Zeng (May), Fei Mi (May), Lifeng Shang (May), Yi R. (May), Fung 5/5/2026

Entropy Centroids as Intrinsic Rewards for Test-Time Scaling

Entropy Centroids: intrinsic reward mechanism for test-time scaling of LLM inference without requiring external reward models.

Ax Boris Shigida, Boris Hanin, Andrey Gromov 5/5/2026

Learning Rate Transfer in Normalized Transformers

Learning rate transfer improvements for Normalized Transformers using alignment exponents for better hyperparameter scaling across model sizes.

Ax Halley Young, Nikolaj Bj\"orner 5/5/2026

Theory Under Construction: Orchestrating Language Models for Research Software Where the Specification Evolves

Methodology for orchestrating LLMs in research software development where specifications evolve, addressing hallucination accumulation failures.

Ax Chenxin Li, Zhengyang Tang, Mingxin Huang, Yunlong Lin, Shijue Huang, Shengyuan Liu, Bowen Ye, Rang Li, Lei Li, Benyou Wang, Yixuan Yuan 5/5/2026

Claw-Eval-Live: A Live Agent Benchmark for Evolving Real-World Workflows

Claw-Eval-Live: live benchmark for workflow agents with refreshable task signals, execution verification, and evolving real-world task evaluation.

Ax Brice Valentin Kok-Shun, Johnny Chan, Gabrielle Peko, David Sundaram 5/5/2026

Agentopic: A Generative AI Agent Workflow for Explainable Topic Modeling

Agentopic: multi-agent LLM workflow for explainable topic modeling with collaborative topic identification, validation, and hierarchical grouping.

Ax Hao Xiao 5/5/2026

From Euler to Dormand-Prince: ODE Solvers for Flow Matching Generative Models

ODE Solvers for Flow Matching: PyTorch implementations and benchmarks of Euler, RK4, and Dormand-Prince solvers for generative model sampling.

Ax Duaa Alim, Mogtaba Alim, Liam Chalcroft 5/5/2026

GAZE: Grounded Agentic Zero-shot Evaluation with Viewer-Level Tools and Literature Retrieval on Rare Brain MRI

GAZE: agentic framework enabling medical vision-language models to iteratively inspect images and retrieve literature via tool-calling and retrieval mechanisms.

Ax Guantian Zheng 5/5/2026

StyleShield: Exposing the Fragility of AIGC Detectors through Continuous Controllable Style Transfer

StyleShield: adversarial attack framework exposing vulnerabilities of AI-generated content detectors through controllable style transfer attacks.

Ax Seyed Mohammad Azimi-Abarghouyi, Mehdi Bennis, Leandros Tassiulas 5/5/2026

Hierarchical Federated Learning for Networked AI: From Communication Saving to Architecture-Aware Design

Hierarchical federated learning architecture for distributed AI systems optimizing communication and computation across multi-tier networks.