AI News

Tracing Stereotypes from Representation to Output in Multilingual LLMs (arxiv.org)

AttnCompress: Dynamic Attention-Guided Trajectory Compression for Software Engineering Agents (arxiv.org)

Exploring Bottom-Up Clustering for Creating Semantic IDs (arxiv.org)

A Measurement Study of LLM Inference Trade-offs Across Edge Continuum Hardware (arxiv.org)

FPicker: Topology-Guided Evolution for Filament Tracing in Low-SNR Microscopy (arxiv.org)

What Eviction Destroys: A Restore-Counterfactual Audit of Forgetting in Agent Memory (arxiv.org)

ACEA: An Adversarial Co-Evolution Arena for Head-to-Head Red-Team and Blue-Team LLM Testing (arxiv.org)

CALIPER: Clean Scenes Cannot Rank Physical Inference in Pretrained Visual Representations (arxiv.org)

X2Streaming-ASR: wait when uncertain, emit when ready for streaming ASR (arxiv.org)

CS-CLIP: Compositional Scene Graph-guided CLIP for Robust Compositional Reasoning (arxiv.org)

3DWay: Generalizing Robot Manipulation via 3D Consistent Waypoints (arxiv.org)

WSPolypNet: Weakly Supervised Polyp Localization in Colonoscopy Videos (arxiv.org)

AI for AI: Optimizing Additional Infrastructure Build-out to Power Artificial Intelligence Data Centers (arxiv.org)

LOBERT: Generative AI Foundation Model for Limit Order Book Messages (arxiv.org)

Dual-Layer Semantic-Spatial Belief Mapping for Aerial Object Goal Navigation (arxiv.org)

IGT @ FinMMEval 2026 Task 2: Question-Type Prompting with Targeted Extraction for Multilingual Financial QA (arxiv.org)

DISEIL: Demonstration Distillation for Sample-Efficient Imitation Learning (arxiv.org)

LLMs for Social Network Modeling: From Network Generation to Dynamic Processes (arxiv.org)

SAFER-Activities: A Dataset for Smart Assessment of Fall Events and Routine Activities (arxiv.org)

Rethinking Sign Language Translation: The Impact of Signer Dependence on Model Evaluation (arxiv.org)

TDDN: Text-aligned Diffused DINO Network for Puzzle Understanding (arxiv.org)

SAFIRE: Safety-Critical Benchmark for Fine-grained Fire and Smoke Understanding in Multimodal LLMs (arxiv.org)

VoT: Vision-of-Thought for Unified Multimodal Representation Alignment (arxiv.org)

Bag of Tricks or Bag of Myths? Reducing Modeling Complexity with Task Knowledge in Explainable Suicide Risk Assessment (arxiv.org)

Do AI Coding Assistants Check Before They Install? A Pre-Registered Demand-Side Audit of Trust Signals in the Research Software Supply Chain (arxiv.org)

TFTrack: A Template-Free Framework for Efficient 3D Point Cloud Tracking (arxiv.org)

An emancipatory vision for designing (generative) AI for learner flourishing (arxiv.org)

Fine PT-PT Web: A High-Quality 41 Billion Tokens Data Collection of the European Portuguese Web (arxiv.org)

Your Agent Says Yes: Interpreting Adversarial Market Behavior Beyond Individual Transactions (arxiv.org)

Quantifying the Engagement Trap: Impact of Short-form Video Recommender Systems on Users with ADHD (arxiv.org)

From Citations to Contributions: LLM-Assisted Credit Scoring of Research Articles (arxiv.org)

Harnessing CLIP and DINO: An Uncertainty-Aware Cascaded Fusion Network for Generalizable Deepfake Image Detection (arxiv.org)

Synergistic Fusion of Topological Structure and Temporal Semantics of Mobility for Urban Region Embedding (arxiv.org)

Constant-Stepsize Stochastic Approximation: Finite-Time Convergence, Gaussian Approximation, and Tail Bounds (arxiv.org)

Climate-ModernBERT: Revisiting Corpus Composition for Domain-Adaptive Continued Pretraining (arxiv.org)

Text Has Curvature (arxiv.org)

Learning functional components of PDEs from data using neural networks (arxiv.org)

ZetaDial: dialing net charge of protein binders at inference time for therapeutic developability (arxiv.org)

Difficulty-Adaptive Tree-Structured Policy Optimization for Expanding Reasoning Coverage in RLVR (arxiv.org)

SUN: Reaching for Novelty in Reinforcement Learning (arxiv.org)

Suan: Rectifying Direct Preference Safety Alignment in Large Language Models (arxiv.org)

Leveraging contextual events on structure-aware next activity prediction (arxiv.org)

Target-Independent Micro-Interventions for Predicting Training Response Across Language-Model Families (arxiv.org)

MoEMB: Scaling Universal Multimodal Embeddings with Efficient Mixture-of-Experts Models (arxiv.org)

Why shared attention vectors fail: a case for outcome-indexed tuning (arxiv.org)

Multi-Level-Set-Based Physics-Driven Neural Network to Solve 3-D Inverse Scattering Problems (arxiv.org)

Leveraging Cardiac Imaging to Improve ECG-Based Detection of Chagas Disease in Resource-Constrained Settings (arxiv.org)

AlphaRJM: Reward-Jump Memory for Stochastic Return-Guided Alpha Discovery (arxiv.org)

Not All Variables Agree: Reliability-Aware Variable-Wise Gradient Surgery for Multivariate Time-Series Forecasting (arxiv.org)

Topological Fraud Detection in Latent Transaction Spaces (arxiv.org)