Ax Florian Geissler, Francesco Carella, Laura Fieback, Jakob Spiegelberg 5/8/2026

Towards Dependable Retrieval-Augmented Generation Using Factual Confidence Prediction

Confidence prediction mechanism for retrieval-augmented generation to assess factuality of retrieved context.

Ax Alan L. McCann 5/8/2026

Governed Metaprogramming for Intelligent Systems: Reclassifying Eval as a Governed Effec

Theory of governed metaprogramming for AI systems that synthesize executable code at runtime with safety constraints.

Ax Yuan-Chi Li, Li-Chi Chen, Sung-Yi Wu, Yu-Che Tsai, Shou-De Lin 5/8/2026

Decision-aware User Simulation Agent for Evaluating Conversational Recommender Systems

LLM-based user simulator for evaluating conversational recommender systems with explicit human decision-making modeling.

Ax Kumar Abhinav 5/8/2026

Career-Aware Resume Tailoring via Multi-Source Retrieval-Augmented Generation with Provenance Tracking: A Case Study

Resume Tailor agentic system using multi-source RAG with career vault for personalized resume suggestions with provenance tracking.

Ax Yuelin Hu, Zhenbo Yu, Zhengxue Cheng, Wei Liu, Li Song 5/8/2026

Maximizing Rollout Informativeness under a Fixed Budget: A Submodular View of Tree Search for Tool-Use Agentic Reinforcement Learning

Framework for selecting informative rollouts in tree search for tool-use agentic RL using submodular optimization under budget constraints.

Ax Kaifeng He, Xiaojun Zhang, Peiliang Cai, Mingwei Liu, Yanlin Wang, Chong Wang, Kaifeng Huang, Bihuan Chen, Xin Peng, Zibin Zheng 5/8/2026

Bridging Generation and Training: A Systematic Review of Quality Issues in LLMs for Code

Systematic review mapping training data quality issues to code generation defects in LLMs, linking generation failures to corpus problems.

Ax Oubo Ma, Ruixiao Lin, Jiahao Chen, Yuan Su, Yong Yang, Shouling Ji 5/8/2026

Shattering the Echo Chamber: Hidden Safeguards in Manuscripts Against the AI Takeover of Peer Review

Study on embedding hidden safeguards in manuscripts to detect when peer reviewers use commercial chatbots for full outsourcing.

Ax Francisco Javier Arceo, Varsha Prasad Narsing 5/8/2026

Securing the Agent: Vendor-Neutral, Multitenant Enterprise Retrieval and Tool Use

Enterprise architecture for securing RAG and agentic AI systems with multi-tenant retrieval, access control, and compliance requirements.

Ax Laurent Guigues 5/8/2026

Graph Normalization: Fast Binarizing Dynamics for Differentiable MWIS

Graph Normalization method as differentiable approximation for NP-hard Maximum Weight Independent Set problem.

Ax Philippe Hansen-Estruch, Jiahui Chen, Vivek Ramanujan, Orr Zohar, Yan Ping, Animesh Sinha, Markos Georgopoulos, Edgar Schoenfeld, Ji Hou, Felix Juefei-Xu, Sriram Vishwanath, Ali Thabet 5/8/2026

ViTok-v2: Scaling Native Resolution Auto-Encoders to 5 Billion Parameters

Scaling Vision Transformer autoencoders to 5B parameters for improved image tokenization across native resolutions.

Ax Junran Wang, Xinjie Shen, Zehao Jin, Pan Li 5/8/2026

How Far Are VLMs from Privacy Awareness in the Physical World? An Empirical Study

Empirical study evaluating privacy awareness of Vision-Language Models deployed as autonomous agents in physical environments.

Ax Faris Chaudhry, Keisuke Yano, Anthea Monod 5/8/2026

Feature Starvation as Geometric Instability in Sparse Autoencoders

Analysis of feature starvation in sparse autoencoders used to interpret LLM representations, proposing geometric stability perspective.

Ax Md Adnan Arefeen, Biplob Debnath, Ravi K. Rajendran, Murugan Sankaradas, Srimat T. Chakradhar 5/8/2026

Open-SAT: LLM-Guided Query Embedding Refinement for Open-Vocabulary Object Retrieval in Satellite Imagery

LLM-guided query embedding refinement system for open-vocabulary object retrieval in satellite imagery using vision-language models.

Ax Lana Do, Shasta Ihorn, Charity M. Pitcher-Cooper, Sanjay Mirani, Gio Jung, Hyunjoo Shim, Zhenzhen Qin, Kien T. Nguyen, Vassilis Athitsos, Ilmi Yoon 5/8/2026

Making AI Drafts Count: A Quality Threshold in Audio Description Workflows

Investigation of how AI-generated draft quality affects human editors' workflow when creating audio descriptions for video accessibility.

Ax Tosin Adewumi, Marcus Liwicki, Foteini Simistira Liwicki, Lama Alkhaled, Hamam Mokayed, Esra S\"umer-Arpak 5/8/2026

Counterargument for Critical Thinking as Judged by AI and Humans

Study evaluating how students use AI-generated counterarguments for critical thinking in writing, examining cheating and cognitive offloading risks.

Ax Rahul Nandakumar, Deepayan Chakrabarti 5/8/2026

COPYCOP: Ownership Verification for Graph Neural Networks

Method to verify ownership of trained Graph Neural Networks by detecting if one GNN was trained to mimic another's embeddings.

Ax Pranav A, Shashank B, Pranav Siddappa, Dominik Seuss, Minal Moharir, Subramanya KN 5/8/2026

Two Steps Are All You Need: Efficient 3D Point Cloud Anomaly Detection with Consistency Models

Efficient 3D point cloud anomaly detection using consistency models instead of diffusion for resource-constrained deployment.

Ax Yllias Chali, Deen Abdullah 5/8/2026

Generating Query-Focused Summarization Datasets from Query-Free Summarization Datasets

Method to automatically generate query keywords from query-free summarization datasets to enable query-focused summarization tasks.

Ax Andrew Zigler 5/8/2026

Mise en Place for Agentic Coding: Deliberate Preparation as Context Engineering Methodology

Methodology for preparing context in AI coding agents using mise en place principles to reduce debugging and improve code quality.

Ax M. Tuluhan Akbulut, Varun Satheesh, Ahmed Jaafar, Alper Ahmetoglu, Shane Parr, Aditya Ganeshan, Shivam Vats, George Konidaris 5/8/2026

Creative Robot Tool Use by Counterfactual Reasoning

Causal reasoning framework enabling robots to creatively use tools beyond their primary purpose via counterfactual simulation.

Ax Yiwei Zhang, Jeremiah Birrell, Reza Ebrahimi, Rouzbeh Behnia, Jason Pacheco, Elisa Bertino 5/8/2026

Information Theoretic Adversarial Training of Large Language Models

Information-theoretic adversarial training approach for improving LLM robustness to adversarial prompts at scale.

Ax Pratik Deshmukh, Atirek Gupta 5/8/2026

On Semantic Loss Fine-Tuning Approach for Preventing Model Collapse in Causal Reasoning

Method using semantic loss to prevent model collapse during transformer fine-tuning on causal reasoning tasks.

Ax Fabrice Harel-Canada, Amit Sahai 5/8/2026

SLAM: Structural Linguistic Activation Marking for Language Models

SLAM: white-box watermarking scheme for LLMs using sparse autoencoders to encode marks in structural geometry without quality loss.

Ax Othmane Kabal, Mounira Harzallah, Fabrice Guillet, Hideaki Takeda, Ryutaro Ichise 5/8/2026

Robustness of Graph Self-Supervised Learning to Real-World Noise: A Case Study on Text-Driven Biomedical Graphs

Study on Graph Self-Supervised Learning robustness to noise in text-derived biomedical knowledge graphs.

Ax Othmane Kabal, Mounira Harzallah, Fabrice Guillet, Hideaki Takeda, Ryutaro Ichise 5/8/2026

A Unified Benchmark for Evaluating Knowledge Graph Construction Methods and Graph Neural Networks

Benchmark for evaluating knowledge graph construction methods and GNN robustness on noisy automatically-extracted graphs.

Ax Raimondo Fanale 5/8/2026

GRALIS: A Unified Canonical Framework for Linear Attribution Methods via Riesz Representation

Unified mathematical framework for linear attribution methods (GradCAM, SHAP, LIME, Integrated Gradients) via Riesz representation theory for interpretability.

Ax Atharva Naik, Yash Mathur, Prakam, Carolyn Rose, David Mortensen 5/8/2026

ReaComp: Compiling LLM Reasoning into Symbolic Solvers for Efficient Program Synthesis

Method compiling LLM reasoning traces into symbolic program synthesizers for efficient program synthesis, achieving 91.3% accuracy on benchmarks without test-time LLM calls.

Ax Shicheng Fan, Nour Elhendawy, Jianle Sun, Ke Fang, Kun Zhang, Yihang Wang, Lu Cheng 5/8/2026

MOSAIC: Module Discovery via Sparse Additive Identifiable Causal Learning for Scientific Time Series

Causal representation learning framework for discovering interpretable modules in scientific time series using sparse additive identifiable methods.

Ax Claire Chen, Jiabao Sean Xiao, Shuze Daniel Liu, Facundo Perez Paolino, Luke Handley, Theophile Jegou du Laz, Ricky Nilsson, Alice Zou, Matthew Graham, Ashish Mahabal 5/8/2026

AstroAlertBench: Evaluating the Accuracy, Reasoning, and Honesty of Multimodal LLMs in Astronomical Classification

Benchmark evaluating multimodal LLMs on specialized astronomical classification tasks, assessing reasoning accuracy and interpretability in scientific applications.

Ax Won-Jun Jang, Si-Hyeon Lee 5/8/2026

Accelerating LMO-Based Optimization via Implicit Gradient Transport

Analysis of LMO-based optimizers like Lion and Muon, proposing implicit gradient transport to accelerate convergence without additional gradient evaluations.

Ax Rixi Xu, Qingyu Liu, Haitao Li, Yushen Chen, Zhikang Niu, Yunting Yang, Jian Zhao, Ke Li, Berrak Sisman, Qinyuan Cheng, Xipeng Qiu, Kai Yu, Xie Chen 5/8/2026

X-Voice: Enabling Everyone to Speak 30 Languages via Zero-Shot Cross-Lingual Voice Cloning

Multilingual zero-shot voice cloning model supporting 30 languages using IPA representation and two-stage training on 420K hours of audio.

Ax Vihaan Nama, Shreya Mendi, Zian Ye, Brinnae Bent 5/8/2026

When2Speak: A Dataset for Temporal Participation and Turn-Taking in Multi-Party Conversations for Large Language Models

Synthetic dataset and pipeline for training LLMs to determine appropriate speaking timing in multi-party conversations, addressing turn-taking calibration.

Ax Xinjie Shen, Rongzhe Wei, Peizhi Niu, Haoyu Wang, Ruihan Wu, Eli Chien, Bo Li, Pin-Yu Chen, Pan Li 5/8/2026

One Turn Too Late: Response-Aware Defense Against Hidden Malicious Intent in Multi-Turn Dialogue

Defense mechanism against multi-turn dialogue attacks where harmful intent is distributed across benign-looking turns in deployed LLMs.

Ax Rose Niousha, Samantha Boatright Smith, Bita Akram, Peter Brusilovsky, Arto Hellas, Juho Leinonen, John DeNero, Narges Norouzi 5/8/2026

The Missing Evaluation Axis: What 10,000 Student Submissions Reveal About AI Tutor Effectiveness

Study analyzing 10,000 student submissions to propose behavioral evaluation metrics for AI tutors beyond pedagogical feedback quality alone.

Ax Dasol Choi, Eugenia Kim, Jaewon Noh, Sang Seo, Eunmi Kim, Myunggyo Oh, Yunjin Park, Brigitta Jesica Kartono, Josef Pichlmeier, Helena Berndt, Sai Krishna Mendu, Glenn Johannes Tungka, \"Ozlem G\"ok\c{c}e, Suresh Gehlot, Katherine Pratt, Amanda Minnich, Haon Park 5/8/2026

XL-SafetyBench: A Country-Grounded Cross-Cultural Benchmark for LLM Safety and Cultural Sensitivity

Multilingual benchmark of 5,500 test cases across 10 countries evaluating LLM safety and cultural sensitivity beyond English-centric approaches.

Ax Dongfang Zhao 5/8/2026

EGA: Adapting Frozen Encoders for Vector Search with Bounded Out-of-Distribution Degradation

Adapter technique for frozen vision encoders in vector search that handles out-of-distribution queries without collapsing performance on unseen classes.

Ax Bing Wang, Ximing Li, Changchun Li, Jinjin Chi, Gang Niu, Masashi Sugiyama 5/8/2026

Decomposing the Basic Abilities of Large Language Models: Mitigating Cross-Task Interference in Multi-Task Instruct-Tuning

Method to reduce cross-task interference in multi-task instruction tuning of LLMs by decomposing basic abilities and mitigating conflicting gradients.

Ax Wesley Hanwen Deng, Mingxi Yan, Sunnie S. Y. Kim, Akshita Jha, Lauren Wilcox, Kenneth Holstein, Motahhare Eslami, Leon A. Gatys 5/8/2026

PersonaTeaming: Supporting Persona-Driven Red-Teaming for Generative AI

Red-teaming methodology for generative AI that incorporates persona-driven approaches to surface diverse risks based on human perspectives and backgrounds.

Ax Naveen Mysore 5/8/2026

Temporal Functional Circuits: From Spline Plots to Faithful Explanations in KAN Forecasting

Framework for interpretable time-series forecasting using Kolmogorov-Arnold Networks with explainable edge functions instead of black-box MLPs.

Ax Bole Ma, Jan Eitzinger, Harald K\"ostler 5/8/2026

Irminsul: MLA-Native Position-Independent Caching for Agentic LLM Serving

Position-independent caching system for agentic LLM serving that addresses cache invalidation issues in multi-turn agent interactions using Multi-Head Latent Attention.

Ax Amrit Nidhi 5/8/2026

Budgeted Attention Allocation: Cost-Conditioned Compute Control for Efficient Transformers

Research on budgeted attention allocation for transformers, enabling multiple cost-quality trade-offs in deployed systems via head-gating mechanisms.

Ax Mohamed Amine Bergach 5/8/2026

When Quantization Is Free: An int4 KV Cache That Outruns fp16 on Apple Silicon

int4 KV cache quantization technique for Apple Silicon that runs faster than fp16 via fused Metal kernels, implemented as HuggingFace Cache subclass.

Ax Lehui Li, Ruixuan Jia, Guo-Ye Yang, Jia Li 5/8/2026

An Empirical Study of Proactive Coding Assistants in Real-World Software Development

Empirical study of proactive LLM-based coding assistants that infer developer intent from IDE interactions and repository context to reduce interaction overhead.

Ax Huchen Yang, Xinghao Dong, Dan Negrut, Jin-Long Wu 5/8/2026

Active Learning for Communication Structure Optimization in LLM-Based Multi-Agent Systems

Active learning approach to optimize communication structure in multi-agent LLM systems, reducing token usage and improving performance with limited training data.

Ax Zhe Liu, Zonghao Ying, Wenxin Zhang, Quanchen Zou, Deyue Zhang, Dongdong Yang, Xiangzheng Zhang, Hao Peng 5/8/2026

SafeHarbor: Hierarchical Memory-Augmented Guardrail for LLM Agent Safety

SafeHarbor: hierarchical memory-augmented safety guardrail for LLM agents that balances security with reducing over-refusal in tool-use scenarios.

Ax Jingjie Ning, Xiaochuan Li, Ji Zeng, Hao Kang, Chenyan Xiong 5/8/2026

Auto Research with Specialist Agents Develops Effective and Non-Trivial Training Recipes

Auto research system using specialist AI agents to autonomously develop training recipes through closed empirical loops with hypothesis testing and code generation.

Ax Md Anwar Hossen, Fatema Siddika, Juan Pablo Munoz, Tanya Roosta, Ali Jannesari 5/8/2026

CRAFT: Forgetting-Aware Intervention-Based Adaptation for Continual Learning

CRAFT: continual learning framework for LLMs that avoids catastrophic forgetting by learning low-rank interventions on hidden representations instead of updating weights.

Ax Mohammad Al Ridhawi, Mahtab Haj Ali, Hussein Al Osman 5/8/2026

Multi-Dimensional Behavioral Evaluation of Agentic Stock Prediction Systems Using LLM Judges with Closed-Loop Reinforcement Learning Feedback

Framework for behavioral evaluation of agentic stock prediction systems using LLM judges and closed-loop RL feedback. Addresses hidden decision quality in aggregated metrics.

Ax Joseph D. Ramsey 5/8/2026

Fourier Feature Methods for Nonlinear Causal Discovery: FFML Scoring and FFCI Testing in Mixed Data

Methods for nonlinear causal discovery using Fourier feature approximations. Scales Gaussian process marginal likelihood scoring and kernel tests.

Ax Haodong Liang, Lifeng Lai 5/8/2026

Transformers Provably Implement In-Context Reinforcement Learning with Policy Improvement

Theoretical research proving transformers can implement policy-improvement reinforcement learning in-context without parameter updates via explicit constructions.