Ax Debosmita Bhaumik, Julian Togelius, Georgios N. Yannakakis, Ahmed Khalifa 5/14/2026

Learning Local Constraints for Reinforcement-Learned Content Generators

Method combining Wave Function Collapse local constraints with reinforcement learning for game content generators balancing visual quality and global properties.

Ax Jie Hu, Changyuan Yan, Yu Zheng, Ziqian Wang, Jiaming Zhang 5/14/2026

Position: Assistive Agents Need Accessibility Alignment

Position paper on accessibility alignment as design objective for assistive agents serving Blind and Visually Impaired users.

Ax Aar\'on L\'opez Garc\'ia, Natalia Criado, Jose Such 5/14/2026

Unweighted ranking for value-based decision making with uncertainty

Fuzzy-unweighted value-based decision framework for alignment of intelligent systems with human values in autonomous decision-making.

Ax Jie Gao, Kaiser Sun, Jen-tse Huang, Katherine Van Koevering, Sijie Ji, Heyuan Huang, Weiyan Shi, Zhuoran Lu, Ziang Xiao, Daniel Khashabi, Mark Dredze 5/14/2026

How to Interpret Agent Behavior

Methods for interpreting autonomous agent behavior by analyzing reasoning trajectories and execution traces from long-running agents like Claude Code.

Ax Yitian Yang, Yiqun Duan, Linghan Huang, Yiqi Zhu, Francesco Bailo, Chunmeizi Su, Huaming Chen 5/14/2026

ScioMind: Cognitively Grounded Multi-Agent Social Simulation with Anchoring-Based Belief Dynamics and Dynamic Profiles

ScioMind framework for LLM-based multi-agent social simulation integrating cognitive grounding with belief dynamics and anchoring effects.

Ax Trung Nguyen Quang, Yiming Gao, Fanyi Pu, Kaichen Zhang, Shuo Sun, Ziwei Liu 5/14/2026

Senses Wide Shut: A Representation-Action Gap in Omnimodal LLMs

IMAVB benchmark identifying representation-action gaps in omnimodal LLMs when textual premises contradict sensory input from video and audio.

Ax Jiayi Zhang, Yongfeng Gu, Jianhao Ruan, Maojia Song, Yiran Peng, Zhiguang Han, Jinyu Xiang, Zhitao Wang, Caiyin Yang, Yixi Ouyang, Bang Liu, Chenglin Wu, Yuyu Luo 5/14/2026

Harnessing Agentic Evolution

Framework combining iterative candidate generation, evaluation, and feedback for agentic evolution, balancing modularity and flexibility in long-horizon search.

Ax Alberto G. Rodr\'iguez Salgado 5/14/2026

History Anchors: How Prior Behavior Steers LLM Decisions Toward Unsafe Actions

Safety study on LLM agents, testing whether models continue harmful actions when prior steps in reasoning logs were harmful using HistoryAnchor-100 benchmark.

Ax S. Akshay, Chaitanya Garg, Ashutosh Gupta, Kuldeep S. Meel, Ajinkya Naik 5/14/2026

Quantifying Sensitivity for Tree Ensembles: A symbolic and compositional approach

Symbolic and compositional approach for verifying sensitivity properties in decision tree ensembles used for safety-critical AI classification tasks.

Ax JC Wu, Norton Lee, Kai Siang Chen 5/14/2026

TokaMind for Power Grid: Cross-Domain Transfer from Fusion Plasma

TokaMind multi-modal transformer foundation model pre-trained on plasma diagnostics, evaluated for cross-domain transfer to industrial and aerospace systems.

Ax Abdul Basit, Saim Rehman, Muhammad Shafique 5/14/2026

Scale-Gest: Scalable Model-Space Synthesis and Runtime Selection for On-Device Gesture Detection

Scale-Gest framework for runtime-adaptive on-device gesture detection on mobile devices with memory and energy constraints.

Ax Siqi Miao, Ziyang Chen, Yuhong Luo, Hans Hao-Hsun Hsu, Mufei Li, Kaiqing Zhang, Pan Li 5/14/2026

Can LLM Agents Simulate Dynamic Networks? A Case Study on Email Networks with Phishing Synthesis

Study evaluating whether LLM multi-agent systems can replicate realistic network dynamics in email/phishing simulations, finding failures in capturing macroscopic topologies.

Ax Haoran Bu, Litian Zhang, Chuxuan Zhang, Zhanyuan Liu, Hui Pang, Xi Zhang 5/14/2026

Beyond Individual Mimicry: Constructing Human-Like Social network with Graph-Augmented LLM Agents

Graph-augmented LLM agents create human-like social networks by coordinating global interactions to evade bot detection.

Ax Saiful Islam Sagor, Tania Haghighi, Minhaj Nur Alam, Erina Baynojir Joyee 5/14/2026

Domain Adaptation of Large Language Models for Polymer-Composite Additive Manufacturing Using Retrieval-Augmented Generation and Fine-Tuning

Domain adaptation of LLMs for additive manufacturing using retrieval-augmented generation and fine-tuning for expert QA.

Ax Mingyeong Kim (Kim Jaechul Graduate School of AI, KAIST), Jungwon Choi (Kim Jaechul Graduate School of AI, KAIST), Chaeyun Jang (Kim Jaechul Graduate School of AI, KAIST), Juho Lee (Kim Jaechul Graduate School of AI, KAIST) 5/14/2026

Bridging the Missing-Modality Gap: Improving Text-Only Calibration of Vision Language Models

Addresses miscalibration in vision-language models when deployed on text-only inputs despite image training.

Ax Liancheng Zhang, Xiaoxi Li, Zhicheng Dou 5/14/2026

TimelineReasoner: Advancing Timeline Summarization with Large Reasoning Models

Applies large reasoning models to timeline summarization with iterative evidence acquisition and event detection.

Ax Kyuyoung Kim, Kevin Wang, Yunfei Xie, Peiyang Xu, Peiyao Sheng, Chen Wei, Zhangyang Wang, Jinwoo Shin, Pramod Viswanath, Sewoong Oh 5/14/2026

Correct Answers from Sound Reasoning: Verifiable Process Supervision for Language Models

Verifiable process supervision post-training framework optimizing both correct answers and sound reasoning in language models.

Ax Yancheng Ling, Zhenlin Qin, Leizhen Wang, Zhenliang Ma 5/14/2026

BoostTaxo: Zero-Shot Taxonomy Induction via Boosting-Style Agentic Reasoning and Constraint-Aware Calibration

Boosting-style LLM framework for zero-shot taxonomy induction using agentic reasoning and constraint-aware calibration.

Ax Dinesh Khandelwal, Gnana Prakash Punnavajhala, GPS Bhargav, Gaurav Pandey, Sachin Joshi, Hima Karanam, Dinesh Raghu 5/14/2026

ToolWeave: Structured Synthesis of Complex Multi-Turn Tool-Calling Dialogues

Framework for synthesizing realistic multi-turn tool-calling dialogues for training agent capabilities. Ensures tools align with meaningful tasks.

Ax Zeyang Zhang, Chengwei Liang, Xingyan Chen, Meiqi Gu, Minrui Luo, Jingzhao Zhang, Tianxing He 5/14/2026

Differences in Text Generated by Diffusion and Autoregressive Language Models

Compares text generation from diffusion vs autoregressive language models. Diffusion models show lower entropy, higher coherence and diversity.

Ax David James Woo, Yangyang Yu, Yilin Huang, Deliang Wang, Kai Guo, Chi Ho Yeung 5/14/2026

Exploring how EFL students talk to and through AI to develop texts

EFL students use prompt engineering and AI negotiation for text development. Analyzes authorship negotiation patterns and writing performance correlation.

Ax Konstantine Arkoudas, Serafim Batzoglou 5/14/2026

Stress-Testing the Reasoning Competence of LLMs With Proofs Under Minimal Formalism

ProofGrid benchmark evaluates LLM reasoning through machine-checkable proofs in minimal formal notation across 15 reasoning tasks.

Ax Zeyu Tang, Sang T. Truong, Deonna Owens, Shreyas Sharma, Yibo Jacky Zhang, Brando Miranda, Sanmi Koyejo 5/14/2026

In-Situ Behavioral Evaluation for LLM Fairness, Not Standardized-Test Scores

Proposes in-situ behavioral evaluation for LLM fairness instead of standardized benchmarks. Shows prompt choices significantly bias fairness scores.

Ax Ivan Letteri 5/14/2026

AgenticAITA: A Proof-Of-Concept About Deliberative Multi-Agent Reasoning for Autonomous Trading Systems

Multi-agent LLM framework for autonomous trading with deliberative reasoning loop replacing traditional algorithmic trading signals.

Ax M. Shamail J. Khan, Nisha L. Raichur, Lucas Heublein, Christian Wielenberg, Alexander Mattick, Tobias Feigl, Christopher Mutschler, Felix Ott 5/14/2026

Active Sensing with Meta-Reinforcement Learning for Emitter Localization from RF Observations

Meta-RL framework for active sensing in GNSS interference localization where agent sequentially explores to infer emitter position.

Ax Chenhao Qiu (Mango TV), Yechao Zhang (Nanyang Technological University), Xin Luo (Mango TV), Shien Song (Mango TV), Xusheng Liu (Mango TV) 5/14/2026

VideoSEAL: Mitigating Evidence Misalignment in Agentic Long Video Understanding by Decoupling Answer Authority

VideoSEAL framework addressing evidence misalignment in agentic long video question answering by decoupling answer authority.

Ax Hongyi Tang, Zhihao Zhu, Yi Yang 5/14/2026

DistractMIA: Black-Box Membership Inference on Vision-Language Models via Semantic Distraction

Black-box membership inference attack on vision-language models using semantic distraction through generated text responses.

Ax Junze Liu, Kun Qian, Florian Dubost, Kai Zhong, Arvind Srinivasan, Nan Chen, Anping Wang, Sam Zhang, Alejandro Mottini, Qingjun Cui, Tian Wang 5/14/2026

3D Primitives are a Spatial Language for VLMs

Study showing 3D geometric primitives as intermediate representation improve VLM spatial reasoning and 3D scene reconstruction from code.

Ax Maryam Amirizaniani, Benjamin Charles Germain Lee, Jevin West, Nicholas Weber 5/14/2026

Training LLMs with Reinforcement Learning for Intent-Aware Personalized Question Answering

RL training approach for personalized question answering in LLMs by explicitly modeling implicit user intent during reasoning.

Ax Nina Corvelo Benz, Eleni Straitouri, Manuel Gomez-Rodriguez 5/14/2026

Learning to Decide with AI Assistance under Human-Alignment

Framework for improving decision-maker trust in AI-assisted predictions through confidence communication and human-alignment mechanisms.

Ax Weichen Yu, Xiaomin Li, Yizhou Zhao, Xiaoze Liu, Ruowang Zhang, Haixin Wang, Yinyi Luo, Chen Henry Wu, Gaurav Mittal, Matt Fredrikson, Yu Hu 5/14/2026

Multi-Rollout On-Policy Distillation via Peer Successes and Failures

On-policy distillation method for LLM training that leverages peer rollout successes and failures to provide dense token-level supervision.

Ax Eun Go, Rohan Deb, Arindam Banerjee 5/14/2026

Plan Before You Trade: Inference-Time Optimization for RL Trading Agents

FPILOT inference-time optimization framework for RL trading agents using model predictive control with price forecasts at deployment.

Ax Nirmal Patel, Fei Wang, Inderjit Dhillon 5/14/2026

ODRPO: Ordinal Decompositions of Discrete Rewards for Robust Policy Optimization

ODRPO method for robust LLM alignment using reinforcement learning from AI feedback with ordinal decomposition of discrete multi-tier rewards.

Ax Yichen Feng, Yuetai Li, Chunjiang Liu, Yuanyuan Chen, Fengqing Jiang, Yue Huang, Hang Hua, Zhengqing Yuan, Kaiyuan Zheng, Luyao Niu, Bhaskar Ramasubramanian, Basel Alomair, Xiangliang Zhang, Misha Sra, Zichen Chen, Radha Poovendran, Zhangchen Xu 5/14/2026

Visual Aesthetic Benchmark: Can Frontier Models Judge Beauty?

Benchmark evaluating whether multimodal LLMs can make accurate aesthetic judgments compared to human expert annotators on image ranking tasks.

Ax Jacqueline L. Mitchell, Chao Wang 5/14/2026

Agentic Interpretation: Lattice-Structured Evidence for LLM-Based Program Analysis

LLM-based program analysis using lattice-structured evidence to enable whole-program analysis beyond static analyzers' capabilities.

Ax Kamel Abdous, Nairouz Mrabah, Mohamed Bouguessa 5/14/2026

Modeling Heterophily in Multiplex Graphs: An Adaptive Approach for Node Classification

Graph neural network method for heterophilic multiplex graphs addressing limitations of homophily-based node classification approaches.

Ax Yifan Chen, Fei Yin, Qingyan Bai, Zicheng Lin, Yujiu Yang 5/14/2026

MMCL-Bench: Multimodal Context Learning from Visual Rules, Procedures, and Evidence

Benchmark for multimodal context learning evaluating models on learning task-local rules from visual and mixed-modality teaching contexts.

Ax Zhiming Yu, Wangtao Lu, Xin Lai 5/14/2026

FePySR: A Neural Feature Extraction Framework for Efficient and Scalable Symbolic Regression

FePySR framework for symbolic regression using neural feature extraction to decompose complex expressions into reusable nonlinear components.

Ax Carmel Baharav, Niclas Boehmer, Bailey Flanigan, Maximilian T. Wittmann 5/14/2026

The End Justifies the Mean: A Linear Ranking Rule for Proportional Sequential Decisions

Democratic design framework for collectively choosing linear ranking decision rules, motivated by AI alignment and participatory design.

Ax Weitai Kang, Xiaohang Zhan, Yizhou Wang, Mang Tik Chiu, Jason Kuen, Kangning Liu, Yan Yan 5/14/2026

Inline Critic Steers Image Editing

Image editing refinement technique using inline critic signals within forward passes to improve heterogeneous editing difficulty across regions.

Ax Boming Liu, Jin Dong, Jamie Lian 5/14/2026

Grid-Orch: An LLM-Powered Orchestrator for Distribution Grid Simulation and Analytics

Grid-Orch uses Model Context Protocol to enable LLM-powered natural language interface for power distribution grid simulation and analysis.

Ax Muhammad Bilal, Jon Crowcroft, Ruizhi Wang, Xiaolong Xu, Schahram Dustdar 5/14/2026

Large Language Models for Agentic NetOps and AIOps: Architectures, Evaluation, and Safety

Survey of LLM applications in network operations and IT operations, covering architectures for incident investigation, root-cause analysis, and self-healing workflows.

Ax Yujia Zheng, Fan Feng, Yuke Li, Shaoan Xie, Kevin Murphy, Kun Zhang 5/14/2026

From Generalist to Specialist Representation

Nonparametric study of learning task-relevant specialist representations from generalist models with identifiability guarantees.

Ax Patrick Callaghan, Reid Simmons, Henny Admoni 5/14/2026

What Do You Think I Think? Accounting for Human Beliefs Using Second-Order Theory of Mind

Agent framework using second-order theory of mind to model and correct human misconceptions about agent knowledge during interaction.

Ax Nirav Diwan, Han Wang, Berkcan Kapusuzoglu, Ramin Moradi, Supriyo Chakraborty, Giri Iyengar, Sambit Sahu, Huan Zhang, Gang Wang 5/14/2026

CoT-Guard: Small Models for Strong Monitoring

Small monitoring models for detecting covert misbehavior in LLM chain-of-thought reasoning, reducing cost vs. large model monitors.

Ax Heejin Do, Shashank Sonkar, Mrinmaya Sachan 5/14/2026

Simulating Students or Sycophantic Problem Solving? On Misconception Faithfulness of LLM Simulators

Framework evaluating whether LLM student simulators faithfully reproduce misconceptions or generate sycophantic responses, important for educational AI tools.

Ax Zhehang Du, Hangfeng He, Weijie Su 5/14/2026

Uncovering Symmetry Transfer in Large Language Models via Layer-Peeled Optimization

Analysis of geometric structure in LLM weights through layer-peeled optimization, studying how cross-entropy loss induces symmetry in learned representations.

Ax Jack Young 5/14/2026

WriteSAE: Sparse Autoencoders for Recurrent State

Sparse autoencoders for decomposing and editing state-space and hybrid recurrent language model caches, supporting interpretability of Mamba and RWKV architectures.

Ax Baris Askin, Muhammed Ustaomeroglu, Anupam Nayak, Gauri Joshi, Guannan Qu, Carlee Joe-Wong 5/14/2026

Emergent and Subliminal Misalignment Through the Lens of Data-Mediated Transfer

Studies emergent misalignment in fine-tuned LLMs as data-mediated transfer phenomenon, analyzing behavioral spillover from harmful fine-tuning.

Ax Shixing Yu, Promit Ghosal, Kyra Gan 5/14/2026

Correcting Influence: Unboxing LLM Outputs with Orthogonal Latent Spaces

Method for attributing LLM predictions to training data tokens using influence functions, targeting healthcare applications requiring explainability.