All published AI Radar articles.

Three new Hugging Face datasets have been released for evaluating Large Language Models (LLMs) on Partial Differential Equation (PDE) problems. These datasets focus on free-generation tasks and include tabular and text modalities.

Three new Hugging Face datasets have been released for evaluating Large Language Models (LLMs) on Partial Differential Equations (PDEs). These datasets, named bermaneh/pde-llm-eval-freegen-xmodal-qwen__qwen3-8-27b, bermaneh/pde-llm-eval-freegen-xmodal-qwen__qwen3-6-27b, and bermaneh/pde-llm-eval-freegen-xmodal-qwen__qwen3-5-27b, contain tabular and text data for free-generation PDE evaluation.

Researchers have introduced G-CARL, a novel reinforcement learning framework designed for patient-oriented medical report interpretation. This framework aims to bridge the gap in existing medical vision-language tasks by addressing the dual requirements of evidence-grounded factuality and context-dependent patient communication.

Researchers introduce 4DAnyone, a framework designed to reconstruct 4D humans from uncalibrated monocular videos. It addresses limitations in existing video diffusion models by generating consistent multi-view videos that are then lifted into 4D Gaussian Splatting (4DGS).

This research introduces a three-agent workflow that integrates conversational data collection, structured data processing, and behavioral prediction for travel behavior analysis. The study evaluates nine locally deployed LLMs, including multimodal configurations, against traditional machine learning models.

Omadia is a self-hostable agentic operating system designed for building, running, and auditing multi-agent AI teams. It supports signed plugins, allows users to bring their own LLM keys, and emphasizes data ownership with EU/GDPR readiness.

A new study compares ceiling-mounted FMCW, IR-UWB, and Wi-Fi radar for in-bedroom human activity and sleep monitoring. The research highlights a trade-off between recognition performance and environmental robustness across these technologies.

DeepSeek has announced the release of DeepSeek-V4-Flash-Vision-Exp. This update pertains to their vision model capabilities.

Researchers have introduced ConceptGuard, a new benchmark designed to evaluate the context-sensitive unlearning capabilities of Large Language Models (LLMs). Existing benchmarks often fail to capture the nuanced requirement of removing harmful knowledge while preserving beneficial information.

Researchers have introduced DreamHand, a framework that repurposes video diffusion models (VDMs) for more robust 3D hand motion recovery from egocentric video. It addresses challenges like object occlusion and hands leaving the frame by using a deterministic geometry encoder.