Ax Fanrui Zhang, Qiang Zhang, Sizhuo Zhou, Jianwen Sun, Chuanhao Li, Jiaxin Ai, Yukang Feng, Yujie Zhang, Wenjie Li, Zizhen Li, Yifan Chang, Jiawei Liu, Kaipeng Zhang 4/6/2026

Code-in-the-Loop Forensics: Agentic Tool Use for Image Forgery Detection

arXiv paper on code-in-the-loop agentic tool use for image forgery detection, unifying low-level artifacts with semantic knowledge from MLLMs.

Ax Sixue Xing, Kerui Wu, Xuanye Xia, Meng Jiang, Jintai Chen, Tianfan Fu 4/6/2026

ClinicalReTrial: Clinical Trial Redesign with Self-Evolving Agents

arXiv paper on ClinicalReTrial, multi-agent system using LLMs to redesign failing clinical trial protocols with actionable recommendations.

Ax Jiayi Yuan, Jonathan N\"other, Natasha Jaques, Goran Radanovi\'c 4/6/2026

AgenticRed: Evolving Agentic Systems for Red-Teaming

arXiv paper on AgenticRed, automated pipeline using in-context learning to evolve red-teaming systems without human-designed workflows.

Ax Bowen Cao, Dongdong Zhang, Yixia Li, Junpeng Liu, Shijue Huang, Chufan Shi, Hongyuan Lu, Yaokang Wu, Guanhua Chen, Wai Lam, Furu Wei 4/6/2026

From Abstract to Contextual: What LLMs Still Cannot Do in Mathematics

arXiv paper analyzing gap between LLM math benchmark performance and real-world application through contextual reasoning benchmark ContextMATH.

Ax A. Humnabadkar, A. Sikdar, B. Cave, H. Zhang, N. Bessis, A. Behera 4/6/2026

From Virtual Environments to Real-World Trials: Emerging Trends in Autonomous Driving

arXiv survey on autonomous driving using synthetic data and virtual environments for training and evaluation.

Ax Yang Li, Yule Liu, Xinlei He, Youjian Zhao, Qi Li, Ke Xu 4/6/2026

Chain-of-Authorization: Embedding authorization into large language models

arXiv paper on embedding authorization mechanisms directly into LLM reasoning to prevent data leakage and unauthorized command execution.

Ax Canfer Akbulut, Rasmi Elasmar, Abhishek Roy, Anthony Payne, Priyanka Suresh, Lujain Ibrahim, Seliem El-Sayed, Charvi Rastogi, Ashyana Kachra, Will Hawkins, Kristian Lum, Laura Weidinger 4/6/2026

Evaluating Language Models for Harmful Manipulation

arXiv paper introducing framework for evaluating harmful AI manipulation through human-AI interaction studies across policy, finance, and health domains.

Ax Zelin Tan, Zhouliang Yu, Bohan Lin, Zijie Geng, Hejia Geng, Yudong Zhang, Mulei Zhang, Yang Chen, Shuyue Hu, Zhenfei Yin, Chen Zhang, Lei Bai 4/6/2026

PAPO: Stabilizing Rubric Integration Training via Decoupled Advantage Normalization

arXiv paper proposing PAPO, integrating process-level evaluation into policy optimization to improve reasoning quality beyond final-answer correctness.

Ax Sha Li, Naren Ramakrishnan 4/6/2026

Experience as a Compass: Multi-agent RAG with Evolving Orchestration and Agent Prompts

arXiv paper on multi-agent RAG with adaptive orchestration and evolving agent prompts to handle complex multi-hop reasoning tasks.

Ax Esakkivel Esakkiraja, Sai Rajeswar, Denis Akhiyarov, Rajagopal Venkatesaramani 4/6/2026

Therefore I am. I Think

Analysis showing LLM reasoning models encode decisions before generating chain-of-thought explanations via linear probes.

Ax Khalid Adnan Alsayed 4/6/2026

When AI Gets it Wrong: Reliability and Risk in AI-Assisted Medication Decision Systems

Study evaluating reliability and risk of AI systems in medication decision-making and healthcare workflows.

Ax Yash Shah, Abhijit Chakraborty, Naresh Kumar Devulapally, Vishnu Lokhande, Vivek Gupta 4/6/2026

OSCAR: Orchestrated Self-verification and Cross-path Refinement

OSCAR framework for mitigating hallucinations in diffusion language models using self-verification during generation.

Ax Tuyen Van Kieu, Chi Linh Hoang, Khanh Van To 4/6/2026

Solving the Two-dimensional single stock size Cutting Stock Problem with SAT and MaxSAT

SAT/MaxSAT framework for solving 2D cutting stock problem in manufacturing optimization.

Ax Sarath Shekkizhar, Romain Cosentino, Adam Earle 4/6/2026

Beyond the Assistant Turn: User Turn Generation as a Probe of Interaction Awareness in Language Models

Research probing whether LLMs encode awareness of conversation continuity by generating user turns after assistant responses.

Ax Thomas Jiralerspong, Xiaoyin Chen, Yash More, Vedant Shah, Yoshua Bengio 4/6/2026

Efficient Causal Graph Discovery Using Large Language Models

Novel framework using LLMs for causal graph discovery via breadth-first search, reducing query complexity from quadratic to linear.

Ax Haoyu Wang, Chunyu Qiang, Tianrui Wang, Cheng Gong, Yu Jiang, Yuheng Lu, Chen Zhang, Longbiao Wang, Jianwu Dang 4/6/2026

Expressive Prompting: Improving Emotion Intensity and Speaker Consistency in Zero-Shot TTS

Improves emotion intensity and speaker consistency in zero-shot LLM-based text-to-speech through expressive prompt design methods.

Ax Jiawei Liu, Fanrui Zhang, Jiaying Zhu, Esther Sun, Dong Li, Qiang Zhang, Zheng-Jun Zha 4/6/2026

ForgeryGPT: A Multimodal LLM for Interpretable Image Forgery Detection and Localization

Multimodal LLM fine-tuned for interpretable image forgery detection and localization providing semantic understanding beyond low-level artifacts.

Ax Yongxiang Liu, Bowen Peng, Li Liu, Xiang Li 4/6/2026

S$^4$ST: A Strong, Self-transferable, faSt, and Simple Scale Transformation for Transferable Targeted Attack

Proposes scale transformation method for transferable targeted adversarial attacks requiring minimal data without surrogate model feedback.

Ax Shin'ya Yamaguchi, Kosuke Nishida, Daiki Chijiwa, Yasutoshi Ida 4/6/2026

Zero-shot Concept Bottleneck Models

Zero-shot concept bottleneck models enabling interpretable predictions without target task training by leveraging zero-shot learning.

Ax Minkyu Choi, S P Sharan, Harsh Goel, Sahil Shah, Sandeep Chinchali 4/6/2026

We'll Fix it in Post: Improving Text-to-Video Generation with Neuro-Symbolic Feedback

Improves text-to-video generation semantic and temporal consistency using neuro-symbolic feedback without retraining the model.

Ax Tianyou Li, Haijun Zou, Jiayuan Wu, Zaiwen Wen 4/6/2026

LMask: Learn to Solve Constrained Routing Problems with Lazy Masking

LMask framework uses dynamic masking with learning to solve constrained routing problems as combinatorial optimization tasks.

Ax Jialin Yang, Dongfu Jiang, Lipeng He, Sherman Siu, Yuxuan Zhang, Disen Liao, Zhuofeng Li, Huaye Zeng, Yiming Jia, Haozhe Wang, Benjamin Schneider, Chi Ruan, Wentao Ma, Zhiheng Lyu, Yifei Wang, Yi Lu, Quy Duc Do, Ziyan Jiang, Ping Nie, Wenhu Chen 4/6/2026

StructEval: Benchmarking LLMs' Capabilities to Generate Structural Outputs

StructEval benchmark systematically evaluates LLM capabilities in generating structured outputs across JSON, HTML, React, SVG and other formats.

Ax Hao Yin, Lijun Gu, Paritosh Parmar, Lin Xu, Tianxiao Guo, Xiujin Liu, Weiwei Fu, Yang Zhang, Tianyou Zheng 4/6/2026

FLEX: A Largescale Multimodal, Multiview Dataset for Learning Structured Representations for Fitness Action Quality Assessment

Introduces FLEX, multimodal multiview dataset for fitness action quality assessment with professional assessment and multiple sensor modalities.

Ax Xingzhong Fan, Hongming Tang, Yue Zeng, M. B. N. Kouwenhoven, Guangquan Zeng 4/6/2026

Category-based Galaxy Image Generation via Diffusion Models

Uses diffusion models for data-driven galaxy image generation without explicit physical parameters, outperforming simulation-based methods.

Ax Vyacheslav Kungurtsev, Monicah Cherop Naibei, Gustav Sir, Akhil Anand, Sebastien Gros, Haozhe Tian, Homayoun Hamedmoghadam 4/6/2026

Mission-Aligned Learning-Informed Control of Autonomous Systems: Formulation and Foundations

Formalizes mission-aligned learning-informed control framework for autonomous physical agents integrating learning with task objectives.

Ax Shaoan Xie, Lingjing Kong, Yujia Zheng, Yu Yao, Zeyu Tang, Eric P. Xing, Guangyi Chen, Kun Zhang 4/6/2026

SmartCLIP: Modular Vision-language Alignment with Identification Guarantees

Proposes modular vision-language alignment architecture improving CLIP's handling of multi-object images and caption misalignment.

Ax Doha Nam, Taehyoun Kim, Duksan Ryu, Jongmoon Baik 4/6/2026

ReDef: Do Code Language Models Truly Understand Code Changes for Just-in-Time Software Defect Prediction?

Introduces ReDef, high-confidence software defect prediction dataset from 22 C/C++ projects, evaluating code language model understanding of changes.

Ax Woojung Song, Dongmin Choi, Yoonah Park, Jongwook Han, Yohan Jo 4/6/2026

Human Psychometric Questionnaires Mischaracterize LLM Psychology: Evidence from Generation Behavior

Compares psychometric questionnaire profiles with actual LLM generation behavior across eight open-source models to assess assessment validity.

Ax Jungeun Lee, Kyungah Lee, Inseok Hwang, SoHyun Park, Young-Ho Kim 4/6/2026

AutiHero: Engaging Parents in Creating Personalized, Multi-path Social Narratives for Autistic Children

GenAI system enabling parents to create personalized multi-path social narratives for autistic children using generative models.

Ax Jason Chen, I-Chun Arthur Liu, Gaurav Sukhatme, Daniel Seita 4/6/2026

ROPA: Synthetic Robot Pose Generation for RGB-D Bimanual Data Augmentation

Generates synthetic robot poses for RGB-D bimanual manipulation data augmentation to improve imitation learning policy training.

Ax Tanise Ceron, Dmitry Nikolaev, Dominik Stammbach, Debora Nozza 4/6/2026

What Is The Political Content in LLMs' Pre- and Post-Training Data?

Analyzes political bias in LLM training data composition across pre and post-training stages to understand sources of model bias.

Ax Zhibo Hou, Zhiyu An, Wan Du 4/6/2026

Beyond Noisy-TVs: Noise-Robust Exploration Via Learning Progress Monitoring

Proposes learning progress monitoring to improve exploration efficiency in reinforcement learning agents when encountering unlearnable noise sources.

Ax Hita Kambhamettu, Alyssa Hwang, Philippe Laban, Andrew Head 4/6/2026

Attribution Gradients: Incrementally Unfolding Citations for Critical Examination of Attributed AI Answers

Introduces attribution gradients technique to improve citation informativeness and evidence transparency in AI answer engines.

Ax Zhongkai Yu, Yue Guan, Zihao Yu, Chenyang Zhou, Zhengding Hu, Shuyi Pei, Yangwook Kang, Yufei Ding, Po-An Tsai 4/6/2026

Patterns behind Chaos: Forecasting Data Movement for Efficient Large-Scale MoE LLM Inference

Forecasts expert selection patterns in Mixture of Experts LLMs to optimize data movement overhead in multi-unit serving systems.

Ax Frank Wu, Mengye Ren 4/6/2026

Local Reinforcement Learning with Action-Conditioned Root Mean Squared Q-Functions

Extends Forward-Forward algorithm to reinforcement learning using action-conditioned Q-functions and layer activity statistics as learning signals.

Ax Federica Bologna, Tiffany Pan, Matthew Wilkens, Yue Guo, Lucy Lu Wang 4/6/2026

CQA-Eval: Designing Reliable Evaluations of Multi-paragraph Clinical QA under Resource Constraints

CQA-Eval evaluation framework for multi-paragraph clinical question answering systems with physician annotations and recommendations for resource-constrained settings.

Ax Subhodip Panda, Dhruv Tarsadiya, Shashwat Sourav, Prathosh A. P, Sai Praneeth Karimireddy 4/6/2026

f-INE: A Hypothesis Testing Framework for Estimating Influence under Training Randomness

f-INE hypothesis testing framework estimates sample influence on model performance while accounting for training randomness, addressing instability in existing influence estimation methods.

Ax Daniel Zhao, Daniel Beaglehole, Taylor Berg-Kirkpatrick, Julian McAuley, Zachary Novack 4/6/2026

Steering Autoregressive Music Generation with Recursive Feature Machines

MusicRFM framework adapts Recursive Feature Machines to enable fine-grained control over frozen pre-trained music generation models via internal activation steering.

Ax Chun-Ming Huang, Li-Heng Chang, I-Hsin Chang, An-Sheng Lee, Hao Kuo-Chen 4/6/2026

Recovering Sub-threshold S-wave Arrivals in Deep Learning Phase Pickers via Shape-Aware Loss

Deep learning approach fixing systematic S-wave detection failures in seismic phase picking via shape-aware loss functions.

Ax Rohit Kundu, Vishal Mohanty, Hao Xiong, Shan Jia, Athula Balachandran, Amit K. Roy-Chowdhury 4/6/2026

SAGA: Source Attribution of Generative AI Videos

SAGA framework for source attribution of AI-generated videos. Identifies specific generative model used instead of binary real/fake detection.

Ax Stefanos Koutoupis, Michaela Areti Zervou, Konstantinos Kontras, Maarten De Vos, Panagiotis Tsakalides, Grigorios Tsagkatakis 4/6/2026

The More, the Merrier: Contrastive Fusion for Higher-Order Multimodal Alignment

Research on contrastive fusion for higher-order multimodal alignment in joint representation learning across multiple modalities.

Ax Jayan Adhikari, Prativa Joshi, Sushish Baral 4/6/2026

Analysis of Invasive Breast Cancer in Mammograms Using YOLO, Explainability, and Domain Adaptation

Deep learning approach using YOLO and ResNet50 for breast cancer detection in mammograms with improved out-of-domain robustness.

Ax Chengqi Dong, Chuhuai Yue, Hang He, Rongge Mao, Fenghe Tang, S Kevin Zhou, Zekun Xu, Xiaohan Wang, Jiajun Chai, Guojun Yin 4/6/2026

Training Multi-Image Vision Agents via End2End Reinforcement Learning

IMAgent: open-source visual agent trained with end-to-end RL for multi-image reasoning tasks, addressing limitations of single-image VLM agents.

Ax Vivek Alumootil, Tuan-Anh Vu 4/6/2026

DePT3R: Joint Dense Point Tracking and 3D Reconstruction of Dynamic Scenes in a Single Forward Pass

Method for dense 3D point tracking and reconstruction in dynamic scenes using single forward pass without requiring known camera poses.

Ax Alessio Buscemi, Tom Deckenbrunnen, Fahria Kabir, Kateryna Mishchenko, Nishat Mowla 4/6/2026

Assessing High-Risk AI Systems under the EU AI Act: From Legal Requirements to Technical Verification

Maps EU AI Act legal requirements to technical verification activities for compliance assessment of high-risk AI systems across member states.

Ax Ziyuan Tao, Chuanzhi Xu, Sandaru Jayawardana, Adnan Mahmood, Wei Bao, Kanchana Thilakarathna, Teng Joon Lim 4/6/2026

FedVideoMAE: Efficient Privacy-Preserving Federated Video Moderation

FedVideoMAE: federated learning framework for privacy-preserving video moderation using self-supervised representations and differential privacy.

Ax Sashuai Zhou, Qiang Zhou, Jijin Hu, Hanqing Yang, Yue Cao, Junpeng Ma, Yinchao Ma, Jun Song, Tiezheng Ge, Cheng Yu, Bo Zheng, Zhou Zhao 4/6/2026

Unified Thinker: A General Reasoning Modular Core for Image Generation

Open-source image generation model with improved reasoning for logic-intensive instruction following, closing gap to closed-source systems.

Ax Honghao Chen, Jiangjie Qiu, Yi Shen Tew, Xiaonan Wang 4/6/2026

Autonomous Computational Catalysis Research via Agentic Systems

Multi-agent framework automating full computational catalysis research lifecycle from conception to publication.

Ax Minghui Chen, Wenlong Deng, James Zou, Han Yu, Xiaoxiao Li 4/6/2026

Textual Equilibrium Propagation for Deep Compound AI Systems

Equilibrium propagation method for optimizing compound AI systems with multiple modules in long-horizon agentic workflows.

Ax J Rosser, Robert Kirk, Edward Grefenstette, Jakob Foerster, Laura Ruis 4/6/2026

Infusion: Shaping Model Behavior by Editing Training Data via Influence Functions

Framework using influence functions to craft training data perturbations inducing targeted model behavior changes.