Ax Zhongxiang Sun, Haolang Lu, Qiang Ma, Qi Li, Qipeng Wang, Liang Pang, Chenyu Liu, Qiankun Li, Hao Sun, Kun Wang, Yi Zeng, Jun Xu, Guoqi Li, Ji-Rong Wen 7/2/2026

NeuroCogMap Reveals Cognitive Organization of Large Language Models

arXiv research introducing NeuroCogMap to reveal cognitive organization within large language models through functional mapping.

Ax Adeel Yousaf, Soumik Ghosh, James Beetham, Amrit Singh Bedi, Mubarak Shah 7/2/2026

The Illusion of High Utility in Safety Alignment of Text-to-Image Diffusion Models

arXiv research showing safety alignment metrics in text-to-image diffusion models may misrepresent utility due to coarse global metrics.

Ax Qiyan Luo, Yingdong Pi, Lekang Wen, Jie Yang, Xiaoyu Wang, Haiming Zhang, Mi Wang 7/2/2026

EO-VGGT: Orbital Ray-Conditioned 3D Foundation Models for Satellite Multi-View Reconstruction

arXiv paper on orbital ray-conditioned 3D foundation models for satellite imagery reconstruction using multi-view data.

Ax Merve Atasever, Cagan Bakirci, Alfredo Reina Corona, Keyan Azbijari, Jyotirmoy V. Deshmukh 7/2/2026

Learning Gait-Aware Quadruped Locomotion with Temporal Logic Specifications

RL framework for quadruped gait learning using Signal Temporal Logic specifications instead of hand-crafted rewards.

Ax Seohyun Lee, Seoung Choi, Dohwan Ko, Jongha Kim, Hyunwoo J. Kim 7/2/2026

VideoSearch-R1: Iterative Video Retrieval and Reasoning via Soft Query Refinement

Iterative video retrieval and reasoning system with soft query refinement for temporal grounding and inter-video analysis.

Ax Ivan Ji (Zihao), Liuyi Hu (Zihao), Harrison (Zihao), Zhao (Xiangjun), Lei Huang (Xiangjun), Qunshu Zhang (Xiangjun), Max (Xiangjun), Fan, Aameek Singh 7/2/2026

Real-Time Hard Negative Sampling via LLM-based Clustering for Large-Scale Two-Tower Retrieval

LLM-based hard negative sampling technique for training two-tower retrieval models in large-scale recommendation systems.

Ax Konstantin Garbers 7/2/2026

Gauging, Measuring, and Controlling Critic Complexity in Actor-Critic Reinforcement Learning

Analysis method for critic model complexity in actor-critic RL using spectral effective-rank entropy.

Ax Xin-Yang Liu, Xiantao Fan, Jian-Xun Wang 7/2/2026

A Multi-Resolution Finite-Volume Inspired Deep Learning Framework for Spatiotemporal Dynamics Prediction

Physics-informed deep learning framework for spatiotemporal dynamics prediction with multi-resolution analysis.

Ax Junlong Liu, Haobo Wang, Weiqi Luo, Xiaojun Jia 7/2/2026

Beyond the Prompt: Jailbreaking Function-Calling LLMs via Simulated Moderation Traces

Security analysis of function-calling LLMs showing jailbreak vulnerabilities through simulated moderation traces.

Ax Anindya Sarkar, Nasik Muhammad Nafi, Isaac Lyngaas, Muralikrishnan Gopalakrishnan Meena, Yevgeniy Vorobeychik 7/2/2026

PAPA: Online Personalized Active Preference Alignment

Active learning framework for personalizing diffusion models via user preference feedback in recommendation systems.

Ax Leyuan Yu, Xiao Tang, Minghao Liu, Xinyuan Li, Xiaokai Bai, Sheng Zhou, Qunshu Lin, Weihao Xuan, Naoto Yokoya 7/2/2026

MindEdit-Bench: Benchmarking Object-Level Counterfactual Spatial Reasoning in VLMs from In-the-Wild Photos

Benchmark for testing spatial counterfactual reasoning in vision-language models using smartphone photo triplets.

Ax Prabod Rathnayaka, Fabian Waschkowski, Lukas Wesemann 7/2/2026

BaseRT: Best-in-Class LLM Inference on Apple Silicon via Native Metal

Native Metal inference runtime for LLMs on Apple Silicon with optimized kernel fusion and memory management.

Ax Trung Thanh Nguyen, Hai Nguyen-Truong, Tu Vo, Hoang M. Truong, Tuan-Anh Vu 7/2/2026

Cross4D-JEPA: Dense Cross-modal Correspondence Distillation for 4D Point Cloud Representation Learning

Self-supervised learning method for 4D point cloud representations using cross-modal distillation from 2D models.

Ax Xuefeng Liu, Mingxuan Cao, Qinan Huang, Thomas Brettin, Rick Stevens, Le Cong 7/2/2026

Active-GRPO: Adaptive Imitation and Self-Improving Reasoning for Molecular Optimization

LLM fine-tuning approach combining imitation learning and RL for multi-step molecular optimization reasoning.

Ax Zhiqi Li, Wen Zhang, Bo Zhu 7/2/2026

Flow-Map GRPO: Reinforcement Learning for Few-Step Flow-Map Generators via Anchored Stochastic Composition

RL method for training few-step diffusion models via policy optimization with stochastic flow-maps.

Ax MD Azizul Hakim, Md Shihab Uddin, Talha Ibne Anis 7/2/2026

Cross-Domain Generalization Failure in Lightweight Intrusion Detection Models for IIoT Networks

Study evaluating cross-domain generalization of lightweight ML models for IIoT intrusion detection systems.

Ax Aiden Durrant, Rahul Baburajan, Georgios Leontidis 7/2/2026

Group-Equivariant Poincar\'e Convolutional Networks

Novel deep learning architecture combining group-equivariance with hyperbolic geometry for visual representation learning.

Ax Srinath Perera, Hasinthaka Piyumal, Frank Leymann, Rania Khalaf 7/2/2026

A Methodology for Investigating AI Patterns Prevalence in Software Repositories

Empirical study of AI design patterns in open-source software repositories to validate their real-world prevalence and utility.

Ax Arya Raeesi, Hanna Roed 7/2/2026

Auditing Forgetting in Limited Memory Language Models

Framework for auditing whether deleted facts persist in limited-memory LLMs through residual parameters or retrieval artifacts.

Ax Boyang Dai, Chaoqi Chen, Yizhou Yu 7/2/2026

Identifying Latent Concepts and Structures for Generalized Category Discovery

Method for discovering novel object categories in open-world settings by identifying latent concepts in vision model representations.

Ax Darshan Patil, Ekaterina Lobacheva, Razvan Pascanu, Sarath Chandar 7/2/2026

Loss Smoothing for Stable Adaptation Under Distribution Shift

ML research on smooth loss transitions during neural network adaptation under distribution shift, improving representation preservation during fine-tuning and RL.

Ax Tobias Christian Nauen, Anosh Billimoria, Federico Raue, Stanislav Frolov, Brian B. Moser, Andreas Dengel 7/2/2026

LUMA: Benchmarking Segmentation via a Lightweight Universal Mask Adapter

Lightweight backbone-agnostic mask adapter for fair comparison of transformer backbones in image segmentation tasks.

Ax Zhao Tian, Yingquan Zhao, Chenyao Suo, Meng Wang, Junjie Chen 7/2/2026

LLVM-Bench: Benchmarking and Advancing Large Language Models for LLVM Compiler Issue Resolution

LLVM-Bench: benchmark with 423 real-world issues for evaluating LLMs on compiler problem resolution and issue triage.

Ax Jaehoon Yoo, Wonjung Kim, Floor Eijkelboom, Chanhyuk Lee, Nicholas M. Boffi, Seunghoon Hong, Jinwoo Kim 7/2/2026

Self-conditioned Flow Map Language Models via Fixed-point Flows

Self-conditioning technique for flow-based language models using fixed-point flows to enhance text generation in few-step generators.

Ax Hanning Yang, Meropi Karakioulaki, Lennart Purucker, Tim Litwin, Cristina Has, Moritz Hess 7/2/2026

LLM-Guided ODE Discovery and Parameter Inference from Small-Cohort Aggregate Data

LLMs guide ODE discovery and parameter inference from small-cohort aggregate data for mechanistic clinical modeling in rare diseases.

Ax Mark Russinovich, Ram Shankar Siva Kumar, Ahmed Salem 7/2/2026

Phantom References: Hallucinated Citations That Survive Peer Review at Top-Tier Conferences

Measures hallucinated citations in peer-reviewed papers generated by LLMs, demonstrating how unsupported references survive conference review.

Ax Lukas Kuhn, Giuseppe Serra, Randall Balestriero, Florian Buettner 7/2/2026

LeVLJEPA: End-to-End Vision-Language Pretraining Without Negatives

Vision-language pretraining without contrastive objectives, adopting non-contrastive methods for dense prediction and frozen visual backbones.

Ax Jalal Mahmud, Eser Kandogan 7/2/2026

Exploring the Semantic Gap in Agentic Data Systems: A Formative Study of Operationalization Failures in Analytical Workflows

Formative study of operationalization failures when LLMs generate queries and construct analytical workflows, identifying semantic gaps in data systems.

Ax Miko{\l}aj S{\l}owikowski, Maciej Witold Majewski 7/2/2026

Recovering Input Text from Hidden States: Study of Gradient-Based Inversion of Decoder-Only Language Models

Studies gradient-based inversion techniques to recover input token sequences from decoder-only LLM hidden states through embedding-space optimization.

Ax Qizhi Jiang, Shuo Wang, Pei Ke, Yuhang Song, Ke Qin 7/2/2026

CAT: Confidence-Adaptive Thinking for Efficient Reasoning of Large Reasoning Models

Adaptive token length reduction for large reasoning models based on confidence estimation to improve inference efficiency.

Ax Aayush Aluru, Chloe Ho, Muhammad Hammouri, Kerry Luo, Myra Malik, Ryan Lagasse, Arjun Bahuguna, Vasu Sharma 7/2/2026

From Personas to Plot: Character-Grounded Multi-Agent Story Generation for Long-Form Narratives

Research paper on MAGNET: Multi-agent framework for long-form narrative generation using persona-grounded character agents.

Ax Midhun Parakkal Unni, Samuel Kaski 7/2/2026

Human-Machine Collaboration on Generative Meta-Learning: Model and Algorithm

Research paper on Generative Meta-Learning with Human Feedback (GMHF). Framework for domain adaptation using expert guidance.

Ax Chengzhi Hu, Xuewen Liu, Jing Zhang, Mengjuan Chen, Zhikai Li, Qingyi Gu 7/2/2026

Post-Training Pruning for Diffusion Transformers

Post-training pruning method for Diffusion Transformers addressing computational overhead in image generation models.

Ax Andrew Bell, George Webber, Andrew P King, Steffen E Petersen, Muhummad Sohaib Nazir, Alistair Young 7/2/2026

Learning Cardiac Motion Priors for Implicit Neural Representations

Learning cardiac motion priors for implicit neural representations to improve motion field estimation and optimization speed.

Ax Ziyue Lin, Xinhang Xie, Kangyi Wang, Siming Chen 7/2/2026

SenseWalk: Agent-Based Semantic Trajectory Simulation Powered by Large Language Models in Zoned Environments

LLM-powered agent simulation for semantic trajectory analysis in zoned environments, modeling human movement with behavioral patterns.

Ax Yaoqi Guo, Yang Liu, Jie M. Zhang, Yun Ma, Yiling Lou, Zhenpeng Chen 7/2/2026

SWE-Doctor: Guiding Software Engineering Agents with Runtime Diagnosis from Multi-Faceted Bug Reproduction Tests

SWE-Doctor guides LLM-based software engineering agents using runtime diagnosis from bug reproduction tests for patch generation.

Ax Aryo Pradipta Gema, Beatrice Alex, Pasquale Minervini 7/2/2026

Logit-Contribution Scoring Identifies Non-Literal Retrieval Heads

Logit-contribution scoring method identifies attention heads in LLMs that synthesize non-literal answers from context in long-context settings.

Ax Iddo Hakim, Sharva Gogawale, Omer Ventura, Gal Grudka, Daria Vasyutinsky-Shapira, Berat Kurar-Barakat, Nachum Dershowitz 7/2/2026

Reading Order Inference for Complex Document Layouts

Graph-based training-free framework for inferring reading order in complex historical document layouts with interleaved text streams.

Ax Hasibur Rahman, Smit Desai 7/2/2026

Behavior-Adaptive Conversational Agents: Toward a Fluid Personality Framework

Framework for LLM-based conversational agents to dynamically adapt personality and persona based on interaction context.

Ax Shaoheng Zhang, Zhichen Li, Jie Mei 7/2/2026

DART-VLN: Test-Time Memory Decay and Anti-Loop Regularization for Discrete Vision-Language Navigation

DART-VLN training-free test-time control framework for discrete vision-language navigation addressing memory staleness and loop avoidance.

Ax Zhishang Xiang, Zerui Chen, Yunbo Tang, Zhimin Wei, Ruqin Ning, Yujie Lin, Qinggang Zhang, Jinsong Su 7/2/2026

MemSyco-Bench: Benchmarking Sycophancy in Agent Memory

MemSyco-Bench benchmarks sycophancy in LLM agent memory systems where retrieved memories cause over-alignment with users.

Ax Jingwei Song, Haofeng Xu, Jie Xiao, Chengke Bao, Jingwei Shi, Pengbin Feng, Weixun Wang, Yuhang Han, Chuan Wu, Linfeng Zhang, Bill Shi 7/2/2026

Staleness-Learning Rate Scaling Laws for Asynchronous RLHF

Analyzes staleness effects in asynchronous RLHF training with stale rollouts, establishes scaling laws for learning rate adjustment.

Ax Arpita Nema, Hanwei Zhu, Xi Zhang, Weisi Lin 7/2/2026

LongVQUBench: Benchmarking Long-Term Video Quality Understanding of Vision-Language Models

LongVQUBench benchmark evaluates vision-language models on long-term video quality assessment with temporal degradation patterns.

Ax James C. Davis, Paschal C. Amusuo, Tanmay Singla, Berk \c{C}akar, Kirsten A. Davis 7/2/2026

Cheap Code, Costly Judgment: A Case Study on Governable Agentic Software Engineering

Study of AI-mediated software engineering focusing on inspection, correction, and maintainability when agents generate code at scale.

Ax Zinan Tang, Yukun Zhang, Shaomian Zheng, Zhuoshi Pan, Qizhi Pei, Dingnan Jin, Jun Zhou, Yujun Wang, Biqing Huang 7/2/2026

CausalMix: Data Mixture as Causal Inference for Language Model Training

CausalMix method treats data mixture optimization as causal inference for LLM training, addressing adaptive data distribution shifts.

Ax Haoran Hao, Shahram Najam Syed, Jeffrey Ichnowski, Jeff Schneider 7/2/2026

FAR: Failure-Aware Retry for Test-Time Recovery and Continual Policy Improvement

FAR framework enables robots to learn from test-time failures and adapt behavior autonomously without human intervention.

Ax Md Abu Hanif Shaikh, Abdullah Al Shafi 7/2/2026

Towards Developing a Multimodal Chat Assistant for University Stakeholders: RAG-based Approach

Multimodal university chatbot using retrieval-augmented generation for domain-specific queries and institutional information access.

Ax Hao Huang 7/2/2026

Muon as a Residual Connection

Mechanistic interpretation of Muon optimizer for neural network training as implicit residual connection.

Ax Bingchen Zhao, Sara Beery, Oisin Mac Aodha 7/2/2026

Autonomous Scientific Discovery via Iterative Meta-Reflection

Autonomous scientific discovery system using iterative meta-reflection for hypothesis generation and validation without predefined research constraints.

Ax Changguo Jia, Tianqi Zhao, Runzhi He, Minghui Zhou 7/2/2026

Skills Are Not Islands: Measuring Dependency and Risk in Agent Skill Supply Chains

Framework for managing dependencies and provenance in LLM agent skill supply chains to prevent duplicated dependencies.