Ax Kristina Schaaff, Quintus Stierstorfer, Valerie Heckel 19d ago

Using AI-based Learning Assistants in Higher Education: A Large-Scale Descriptive Analysis

Large-scale study of AI-based learning assistant usage patterns across 77,543 students in distance education.

Ax Yifan Zhou, Qihao Yang, Yan Li, Donggang Li, Xiru Hu, Hokin Deng, Ziyang Gong, Xuanyi Zhou, Huacan Wang, Xiangchao Yan, Wanghan Xu, Wenlong Zhang, Shaofeng Zhang, Yue Zhou, Yifan Yang, Zhihang Zhong, Xue Yang 19d ago

Ideas Have Genomes: Benchmarking Scientific Lineage Reasoning and Lineage-Grounded Idea Generation

Benchmark for evaluating LLMs on scientific lineage reasoning and idea generation grounded in paper citation inheritance structures.

Ax Oded Ovadia, Eli Turkel 19d ago

LLT: Local Linear Transformer for PDE Operator Learning

Local linear transformer architecture for PDE operator learning with reduced computational complexity and local interaction bias.

Ax Wentao Lu 19d ago

ReCoLoRA: Spectrum-Aware Recursive Consolidation for Continual LLM Fine-Tuning

Spectrum-aware framework for continual LLM fine-tuning using recursive consolidation of LoRA adapters across task sequences.

Ax J. de Curt\`o, I. de Zarz\`a 19d ago

Collective Intelligence with Foundation Models

Multi-agent framework coordinating solver, critic, and aggregator agents for consensus reasoning with semantic and procedural evaluation.

Ax Haozhan Tang, Zerui Wang, Yuxian Gu, Song Han, Han Cai 19d ago

Jet-Long: Efficient Long-Context Extension with Dynamic Bifocal RoPE

Dynamic bifocal RoPE method for efficient zero-shot context extension in LLMs, supporting long-context agentic workflows and RAG.

Ax Hari Prasad 19d ago

A Transdiagnostic Space of Disorder Like Phenotypes in Reinforcement Learning Agents

Computational psychiatry study modeling psychological disorders in RL agents through controllable manipulation of cognitive appraisal signals.

Ax Ezgi Korkmaz 19d ago

Principled Analysis of Deep Reinforcement Learning Evaluation and Design Paradigms

Analysis of deep reinforcement learning evaluation paradigms and design principles from foundational DQN to recent algorithmic advances.

Ax Eric Jiang, Xiao Liang, Yikai Zhang, Yingjia Wan, Mengting Li, Haikang Deng, Alexander K. Taylor, Justin Baker, Rushil Raghavan, Junyi Zhang, Ying Nian Wu, Andrea L. Bertozzi, Kai-Wei Chang, Raghu Meka, Matthew Sottile, Nanyun Peng, Amit Sahai, Terence Tao, Wei Wang 19d ago

From Solvers to Research: Large Language Model-Driven Formal Mathematics at the Research Frontier

LLM-driven formal mathematics system for frontier research using interactive theorem proving, advancing beyond well-defined problems.

Ax Jingyao Cai, Shuaijun Liu, Abdul Rehman, Yutong Guo, Qin Tian, Thomas Dolby, Sue Green, Chantel Cox, Xiaosong Yang 19d ago

From Triggers to Emotions: A CPM-Grounded Appraisal Multi-Agent for Dynamic Emotional Evolution in Persona-Based Dialogue

Multi-agent LLM framework for dynamic emotional evolution in persona-based dialogue using cognitive appraisal models.

Ax Sirui Lu, Erickson Tjoa, J. Ignacio Cirac 19d ago

Multi-agent Autoformalization of Tensor Network Theory

Multi-agent LLM system for autonomous formalization of tensor network theory research, coordinated through structured blueprints with periodic human review.

Ax Erik Jagnandan, Mulugeta Haile, Gregory Barber, Pratik Chaudhari 19d ago

Time-to-Collision Based Dynamic Obstacle Avoidance Using Pretrained Vision Models for Robots in Unstructured Environments

Data-efficient vision-based obstacle avoidance for robots using pretrained vision models, avoiding sim-to-real transfer problems.

Ax Anupam Wagle, Ifrat Ikhtear Uddin, Chaowei Zhang, Longwei Wang 19d ago

Mechanistic Interpretability of LLM Jailbreaks via Internal Attribution Graphs

Research using internal attribution graphs to understand LLM jailbreak mechanisms and how adversarial perturbations alter internal reasoning.

Ax Nobin Sarwar, Shubhashis Roy Dipta, Zheyuan Liu, Vaidehi Patil 19d ago

Multimodal Unlearning Across Vision, Language, Video, and Audio: Survey of Methods, Datasets, and Benchmarks

Survey of multimodal unlearning methods, datasets, and benchmarks for VLMs, DMs, LLMs addressing sensitive/biased training data removal.

Ax Mohamed Amine Merzouk, Nolan Smyth, Damiano Fornasiere, Linh Le, David Williams-King, Adam Oberman 19d ago

Efficient Safety Alignment of Language Models via Latent Personality Traits

Latent Personality Alignment method for efficient safety alignment of LLMs using adversarial training on 66 harm-agnostic statements.

Ax Giulia Marchiori Pietrosanti, Giulio Rossolini, Giorgio Buttazzo 19d ago

Adversarial Decoys: Misdirecting Attention-Based Defenses in ViT

Research on adversarial decoys attacking Vision Transformers by misdirecting attention-based defenses. Security and robustness study.

Ax Claudio Meggio, Johan Pensar, Riccardo De Bin 19d ago

path_boost: A Python Package for Interpretable Graph-Level Prediction using Path-Based Gradient Boosting

path_boost Python package for interpretable graph-level prediction using path-based gradient boosting algorithm. Open source tool.

Ax Tommaso Cerruti, Tim Rieder, George Rowlands, Lingfeng Jin, Imanol Schlag 19d ago

Linear Attention Architectures: Mechanisms, Trade-offs, and Cross-Layer Routing

Comparative study of linear attention architectures (DeltaNet, Gated DeltaNet, Kimi Delta) addressing quadratic cost limitations of softmax attention.

Ax Yihong Xu, Mingyu Kang, Linyuan L\"u 19d ago

A Multi-cluster Boundary Learning Method for Out-of-Scope Intent Detection via MiniLM Embedding

Research on out-of-scope intent detection using multi-cluster boundary learning with MiniLM embeddings for human-machine interaction systems.

Ax Xiuyi Lou, Zicheng Xu, Yu-Neng Chuang, Hoang Anh Duy Le, Zhaozhuo Xu, Guanchu Wang, Vladimir Braverman 19d ago

When Implausible Tokens Get Reinforced: Tail-Aware Credit Calibration for LLM Reinforcement Learning

Tail-aware credit calibration method for LLM reinforcement learning that addresses positive-credit contamination in uniform token advantage assignment.

Ax Shyam Agarwal, Courtney Miller, Christian K\"astner, Bogdan Vasilescu 19d ago

3100 Opinions on Code Review in an AI World: Building Causal Theory from Practitioner Discourse

Qualitative analysis of 3100 practitioner opinions on code review in context of AI coding agents using causal inference.

Ax A. Sayyad, J. Emmons, S. Jones, T. Lin, H. Krishnan 19d ago

A Reliability Assessment of LALM Audio Judges for Full-Duplex Voice Agents

Empirical reliability assessment of Gemini models as audio judges for full-duplex voice agents, validated against human raters.

Ax Yufei Xia, Anjun Gao, Yueyang Quan, Zhuqing Liu, Minghong Fang 19d ago

Who Broke the System? Failure Localization in LLM-Based Multi-Agent Systems

Failure localization framework for diagnosing which agent caused system-level failures in LLM-based multi-agent systems.

Ax Anjun Gao, Yueyang Quan, Zhuqing Liu, Minghong Fang 19d ago

Beware What You Autocomplete: Forensic Attribution of Backdoored Code Completions

CodeTracer forensic framework for detecting and attributing backdoor attacks in code completion models.

Ax Nivasini Ananthakrishnan, Mark Bedaywi, Michael I. Jordan, Stuart Russell, Nika Haghtalab 19d ago

Provably Optimal Learning Algorithms for Assistance Games

Provably efficient learning algorithms for repeated assistance games where informed and uninformed agents optimize shared rewards.

Ax Riccardo Revalor, Jalees Rehman, Debjit Pal 19d ago

Can We Trust LLM's Logic? Quantifying Uncertainty, Coherence, and Robustness via a Graph-Based Framework

Graph-based framework to quantify uncertainty and coherence in LLM reasoning chains beyond final-answer agreement.

Ax Emily Jin, Joy Hsu, Yiqing Xu, Weiyu Liu, Nick Haber, Jiajun Wu 19d ago

APIVOT: Adaptive Planning with Interleaved Vision-Language Thoughts

Vision-language model planner for long-horizon robot tasks that interleaves semantic reasoning and geometric constraint checking adaptively.

Ax Ryota Kobayashi, Tsubasa Hirakawa, Takayoshi Yamashita, Hironobu Fujiyoshi, Yasunori Ishii, Tomoyuki Okuno, Kazuki Kozuka 19d ago

Structured Pruning of Large Language Models via Power Transformation and Sign-Preserving Score Aggregation with Adaptive Feature Retention

Structured pruning method for LLMs using power transformation and sign-preserving score aggregation with adaptive feature retention.

Ax Shuang Wang, Chenxu Wang, Hantong Xing, Hanlin Mo, Lirong Han, Licheng Jiao 19d ago

DKDNet: Dual Knowledge and Data-Driven Network for Cross-Domain Automatic Modulation Classification

Deep learning method for automatic modulation classification with domain adaptation using knowledge and data-driven approaches.

Ax Dhruv Agarwal, Anya Shukla, Tanya Goyal, Aditya Vashistha 19d ago

PLURAL: A Global Dataset for Value Alignment

PLURAL dataset with 92 countries of value-aligned preference data from Integrated Values Survey to reduce Western bias in LLMs.

Ax Kshitij Dani, Cordero Core, Landung Setiawan, Carlos Garcia Jurado Suarez, Anshul Tambay, Vani Mandava, Anant Mittal 19d ago

Aleena: Alignment Agent for Research Software Engineering Collaborations

AI agent for research software engineering that maintains alignment across distributed artifacts like meetings, pull requests, and GitHub issues.

Ax Rapha\"el Sarfati, Pratyush Ranjan Tiwari, Siddharth Boppana, Christopher J. Earls, Srikar Varadaraj, Eric Ho 19d ago

What LLM Forecasters Know but Don't Say: Probing Internal Representations for Calibration and Faithfulness

Probing internal LLM representations to improve calibration and faithfulness of forecasting models beyond chain-of-thought outputs.

Ax Samuel Tetteh, Udip Shrestha, Joshua R. Waite, Cody Fleming 19d ago

Who Analyses the Analyser? Self-Validating LLM Hazard Analysis with Constitutional Meta-STPA

Framework for self-validating LLM-assisted safety analysis using Constitutional AI and Systems-Theoretic Process Analysis to detect hallucinations.

Ax Yidong Ouyang, Zhe Wang, Sourav Bhabesh, Dmitriy Bespalov 19d ago

Reinforcing the Generation Order of Multimodal Masked Diffusion Models

Investigation of adaptive token generation ordering in diffusion language models for text-to-image synthesis and multimodal understanding tasks.

Ax Jiantong Jiang, Peiyu Yang, Rui Zhang, Feng Liu 19d ago

Towards Efficient Large Language Model Serving: A Survey on System-Aware KV Cache Optimization

Survey on KV cache optimization techniques for efficient LLM serving, focusing on system-aware infrastructure for reducing memory and latency.

Ax Mayank Singal 19d ago

When Thinking Hurts: Epistemic Signals in the Reasoning Chains of Visual Language Models

Empirical analysis of uncertainty quantification in vision language models with chain-of-thought reasoning, testing four models on adversarial samples.

Ax Maud Ehrmann, Emanuela Boros, Juri Opitz, Andrianos Michail, Florian Wagner, Simon Clematide 19d ago

ICDAR 2026 HIPE-OCRepair Competition on LLM-Assisted OCR Post-Correction for Historical Documents

ICDAR competition on LLM-assisted OCR post-correction for historical documents. Evaluates LLM effectiveness across languages and document types.

Ax Corban Villa, Alp Eren Ozdarendeli, Sijun Tan, Raluca Ada Popa 19d ago

Prismata: Confining Cross-Site Prompt Injection in Web Agents

Security framework preventing prompt injection attacks on web-based autonomous agents. Confines cross-site injection by separating task from untrusted content.

Ax Xuefei Wang 19d ago

Out of Sight: Compression-Aware Content Protection against Agentic Crawlers

Defense against LLM-based agentic crawlers exploiting context compression. Revisits agent pipelines to identify new threat surface.

Ax Qi Lyu, Baicheng Liu, Xudong Wang, Jiahua Dong, Lianqing Liu, Zhi Han 19d ago

LEEVLA: Seeing What Matters in Latent Environment Evolution for Vision-Language-Action

Vision-language-action model for robotics emphasizing task-critical visual evidence. Handles complex dynamic scenarios with latent environment modeling.

Ax Lorenzo Pant\`e, Andrea Fanti, Roberto Capobianco 19d ago

Open-ended Multi-agent Autocurricula via Visual Inspection of Policies with Multi-modal LLMs

Open-ended RL curriculum using multi-modal LLMs to visually inspect agent policies and assess task difficulty. Enables complex skill development.

Ax Hugo Garc\'ia Cuesta, Pablo Mateo Torrej\'on, Alfonso S\'anchez-Maci\'an 19d ago

Multi-Agent Firewall Architecture for Privacy Protection of Sensitive Data in Interactions with Language Models

Open-source privacy firewall for LLM interactions combining browser extension and proxy. Intercepts HTTP/S and WebSocket traffic for data protection.

Ax Lea Roxanne Muth, Marian Margraf 19d ago

From Legacy Documentation to OSCAL: An MCP-Based Agent Pipeline for Threat-Informed Continuous Compliance in Critical Infrastructure

Multi-agent pipeline using MCP for converting legacy documentation to NIST OSCAL compliance format. Non-invasive threat assessment for critical infrastructure.

Ax Weiming Sheng, Jinlang Wang, Manuel Barros, Aldrin Montana, Jacopo Tagliabue, Luca Bigon 19d ago

GitLake: Git-for-data for the agentic lakehouse

Git-based versioning system for data lakehouse enabling agents to work on isolated branches with human review. Production system with atomic publishing.

Ax Giuliano Gorgone, Fausto Carcassi 19d ago

TypeProbe: Recovering Type Representations from Hidden States of Pre-trained Code Models

Analyzes type information encoding in pre-trained code models through hidden state probing. Shows cross-lingual type representations emerge from untyped code.

Ax Xueke Zhu, Qingyan Meng, Liutao Yu, Wei Zhang, Zhengyu Ma, Huihui Zhou, Yonghong Tian 19d ago

FSD-VLN: Fast-Slow Dual-System Modeling for Aerial Long-Horizon Vision-Language Navigation

Fast-Slow dual-system for UAV vision-language navigation balancing semantic reasoning with real-time flight control.

Ax M\'at\'e Gedeon, P\'eter Mihajlik 19d ago

On the Role of Conversational Timing in Synthetic Training Data for ASR

Studies timing properties in synthetic conversational speech data as controllable variable for training conversational ASR systems.

Ax Matthias Wei{\ss}, Athreya Hosahalli Prakash, Maurice Artelt, Falk Dettinger, Nasser Jazdi, Michael Weyrich 19d ago

Self-Adaptive Anomaly Detection with Reinforcement Learning and Human Feedback in Connected Vehicles

Self-adaptive anomaly detection system using reinforcement learning and human feedback for connected vehicle monitoring.

Ax Jing Jie Tan, Ban-Hoe Kwan, Danny Wee-Kiat Ng, Yan-Chai Hum, Shih-Yu Lo, Po-An Chen, Noriyuki Kawarazaki, Kosuke Takano, Anissa Mokraoui 19d ago

Large-Language-Models-as-a-Judge in Theory-Agnostic Adaptive Metric-Alignment for Prototypical Networks in Personality Recognition

Uses LLMs as judges to align metric learning in personality recognition, moving beyond theory-dependent taxonomies.

Ax Xuerun Yan, Zhexi Lian, Nuoheng Zhang, Shiyu Fang, Haoran Wang, Chen Lv, Jia Hu, Binyang Song 19d ago

WCog-VLA: A Dual-Level World-Cognitive Vision-Language-Action Model for End-to-End Autonomous Driving

Dual-level Vision-Language-Action model combining semantic forecasting and world evolution for proactive autonomous driving.