2/24/2026
Next Reply Prediction X Dataset: Linguistic Discrepancies in Naively Generated Content
Dataset and analysis of linguistic discrepancies when LLMs generate content naively for social science research proxies.
2/24/2026
FUSAR-GPT : A Spatiotemporal Feature-Embedded and Two-Stage Decoupled Visual Language Model for SAR Imagery
FUSAR-GPT visual language model for SAR imagery interpretation with spatiotemporal feature embedding.
2/24/2026
Visual Prompt Guided Unified Pushing Policy
Unified pushing policy for robotic manipulation using visual prompting with broader applicability across scenarios.
2/24/2026
HybridFL: A Federated Learning Approach for Financial Crime Detection
HybridFL framework for federated learning addressing hybrid data distribution in financial crime detection.
2/24/2026
How to Allocate, How to Learn? Dynamic Rollout Allocation and Advantage Modulation for Policy Optimization
Dynamic rollout allocation and advantage modulation for LLM reasoning with verifiable rewards, improving policy optimization.
2/24/2026
Evaluating SAP RPT-1 for Enterprise Business Process Prediction: In-Context Learning vs. Traditional Machine Learning on Structured SAP Data
Independent evaluation of SAP RPT-1 tabular foundation model vs XGBoost/LightGBM on structured enterprise data.
2/24/2026
Scaling Laws for Precision in High-Dimensional Linear Regression
Theoretical analysis of how quantization affects model and data capacities in low-precision high-dimensional linear regression.
2/24/2026
No Need For Real Anomaly: MLLM Empowered Zero-Shot Video Anomaly Detection
LAVIDA uses multimodal LLMs for zero-shot video anomaly detection without labeled anomaly data.
2/24/2026
DGPO: RL-Steered Graph Diffusion for Neural Architecture Generation
DGPO combines RL fine-tuning with graph diffusion for neural architecture search on directed acyclic graphs.
2/24/2026
CORVET: A CORDIC-Powered, Resource-Frugal Mixed-Precision Vector Processing Engine for High-Throughput AIoT applications
CORDIC-based vector processing engine for edge AI with runtime-adaptive mixed-precision for IoT applications.
2/24/2026
Taming Preconditioner Drift: Unlocking the Potential of Second-Order Optimizers for Federated Learning on Non-IID Data
Analysis and solution for preconditioner drift in federated second-order optimizers on non-IID data, improving convergence stability.
2/24/2026
Safe and Interpretable Multimodal Path Planning for Multi-Agent Cooperation
Multimodal path planning for multi-agent cooperation using language communication for safe decentralized coordination.
2/24/2026
TOPReward: Token Probabilities as Hidden Zero-Shot Rewards for Robotics
TOPReward uses token probabilities from VLA models as zero-shot reward signals for robot RL training with improved sample efficiency.
2/24/2026
Learning to Reason for Multi-Step Retrieval of Personal Context in Personalized Question Answering
Multi-step retrieval method for personalized QA using reasoning to extract relevant personal context beyond direct query-based retrieval.
2/24/2026
Anatomy of Agentic Memory: Taxonomy and Empirical Analysis of Evaluation and System Limitations
Taxonomy and empirical analysis of memory systems in LLM agents, evaluating benchmarks, metrics, and performance across model backbones for long-horizon reasoning.
2/24/2026
City Editing: Hierarchical Agentic Execution for Dependency-Aware Urban Geospatial Modification
Hierarchical agentic system for automated urban geospatial modification using multimodal reasoning to handle interdependent planning changes.
2/24/2026
Soft Sequence Policy Optimization: Bridging GMPO and SAPO
Policy optimization method bridging GMPO and SAPO using sequence-level importance sampling for improved LLM alignment training.
2/24/2026
Smooth Gate Functions for Soft Advantage Policy Optimization
Soft Adaptive Policy Optimization with smooth gate functions replacing hard clipping for more stable LLM training compared to GRPO.
2/24/2026
Active perception and disentangled representations allow continual, episodic zero and few-shot learning
Framework combining active perception and disentangled representations for continual and few-shot learning without destructive interference.
2/24/2026
Stable Deep Reinforcement Learning via Isotropic Gaussian Representations
Theoretical analysis and empirical validation of isotropic Gaussian embeddings for stable deep reinforcement learning under non-stationary environments.
2/24/2026
Hilbert-Augmented Reinforcement Learning for Scalable Multi-Robot Coverage and Exploration
Multi-robot coverage framework integrating Hilbert space-filling curves into DQN and PPO for scalable decentralized exploration and learning.
2/24/2026
When AI Teammates Meet Code Review: Collaboration Signals Shaping the Integration of Agent-Authored Pull Requests
Empirical study of autonomous coding agents' pull requests on GitHub, analyzing integration outcomes and review collaboration signals using AIDev dataset.
2/24/2026
Red-Teaming Claude Opus and ChatGPT-based Security Advisors for Trusted Execution Environments
Red-team evaluation of Claude Opus and ChatGPT as security advisors for trusted execution environments, identifying LLM vulnerabilities in security guidance.
2/24/2026
SenTSR-Bench: Thinking with Injected Knowledge for Time-Series Reasoning
Benchmark for time-series reasoning combining general LLM reasoning with domain-specific time-series knowledge through injected knowledge approach.
2/24/2026
Can Large Language Models Replace Human Coders? Introducing ContentBench
ContentBench benchmark suite measuring how well low-cost LLMs perform interpretive coding tasks versus humans, tracking cost-accuracy tradeoffs.
2/24/2026
Scale-PINN: Learning Efficient Physics-Informed Neural Networks Through Sequential Correction
Sequential correction algorithm for training physics-informed neural networks more efficiently on partial differential equations.
2/24/2026
Making Conformal Predictors Robust in Healthcare Settings: a Case Study on EEG Classification
Evaluates conformal prediction methods for robust uncertainty quantification in EEG clinical diagnosis under distribution shifts.
2/24/2026
Federated Learning Playground
Interactive browser-based platform for learning federated learning concepts with real-time visualization of heterogeneous data distributions and aggregation algorithms.
2/24/2026
Botson: An Accessible and Low-Cost Platform for Social Robotics Research
Open-source anthropomorphic social robot platform powered by LLMs for accessible AI research in human-robot interaction.
2/24/2026
Pyramid MoA: A Probabilistic Framework for Cost-Optimized Anytime Inference
Hierarchical mixture-of-agents architecture for cost-optimized LLM inference routing between small and large models.
2/24/2026
Large Language Model-Assisted UAV Operations and Communications: A Multifaceted Survey and Tutorial
Survey on integrating large language models into UAV systems for environmental understanding, swarm coordination, and task reasoning.
2/24/2026
Cost-Aware Diffusion Active Search
Active search algorithm for autonomous agents balancing exploration-exploitation with cost-aware decision making.
2/24/2026
Agentic AI as a Cybersecurity Attack Surface: Threats, Exploits, and Defenses in Runtime Supply Chains
Security analysis of LLM-based agentic systems examining runtime attack surfaces, tool invocation risks, and exploitation vectors.
2/24/2026
CTC-TTS: LLM-based dual-streaming text-to-speech with CTC alignment
LLM-based text-to-speech system using CTC alignment for low-latency dual-streaming synthesis with improved training sequences.
2/24/2026
Interpolation-Driven Machine Learning Approaches for Plume Shine Dose Estimation: A Comparison of XGBoost, Random Forest, and TabNet
Comparison of ML approaches (XGBoost, Random Forest, TabNet) for radiation dose estimation in nuclear safety. Physics simulation.
2/24/2026
Tri-Subspaces Disentanglement for Multimodal Sentiment Analysis
Multimodal sentiment analysis approach using tri-subspace disentanglement for pairwise modality-shared signals.
2/24/2026
Detecting High-Potential SMEs with Heterogeneous Graph Neural Networks
Heterogeneous Graph Transformer predicting high-potential SMEs from public SBIR funding data. Business intelligence application.
2/24/2026
CLCR: Cross-Level Semantic Collaborative Representation for Multimodal Learning
Multimodal learning method addressing semantic misalignment via cross-level collaborative representation fusion.
2/24/2026
Satellite-Based Detection of Looted Archaeological Sites Using Machine Learning
Machine learning pipeline using satellite imagery to detect looted archaeological sites in Afghanistan. Application of computer vision.
2/24/2026
VecFormer: Towards Efficient and Generalizable Graph Transformer with Graph Token Attention
Graph Transformer variant with token-level attention for improved scalability and out-of-distribution generalization on large graphs.
2/24/2026
PedaCo-Gen: Scaffolding Pedagogical Agency in Human-AI Collaborative Video Authoring
Pedagogically-informed human-AI system for collaborative instructional video generation using cognitive learning theory.
2/24/2026
Localized Concept Erasure in Text-to-Image Diffusion Models via High-Level Representation Misdirection
Concept erasure technique for text-to-image diffusion models to prevent harmful content generation via representation misdirection.
2/24/2026
Compositional Planning with Jumpy World Models
Compositional planning approach using pre-trained policy composition for temporal abstraction in agent decision-making.
2/24/2026
Denoising Particle Filters: Learning State Estimation with Single-Step Objectives
Particle filtering algorithm for state estimation trained with single-step objectives instead of sequence unrolling. Robotics application.
2/24/2026
Representation Stability in a Minimal Continual Learning Agent
Study of representational dynamics in minimal continual learning agent with persistent state across executions.
2/24/2026
Carbon-Aware Governance Gates: An Architecture for Sustainable GenAI Development
Architecture for embedding carbon-aware governance gates in GenAI-assisted software development workflows.
2/24/2026
Hexagon-MLIR: An AI Compilation Stack For Qualcomm's Neural Processing Units (NPUs)
Open-source MLIR-based compilation stack for deploying Triton kernels and PyTorch models on Qualcomm NPUs.
2/24/2026
SafePickle: Robust and Generic ML Detection of Malicious Pickle-based ML Models
ML-based detection of malicious pickle-serialized models in repositories like Hugging Face to prevent RCE attacks.
2/24/2026
MAS-FIRE: Fault Injection and Reliability Evaluation for LLM-Based Multi-Agent Systems
Fault injection framework (MAS-FIRE) for evaluating reliability of LLM-based multi-agent systems under semantic failures.
2/24/2026
LLM-enabled Applications Require System-Level Threat Monitoring
System-level threat monitoring and reliability challenges for LLM-enabled applications in production environments.