…Generated by Qwen/Qwen2.5-Coder-32B-Instruct While video streaming understanding has made significant strides, real-world applications, such as live sports broadcasting, autonomous driving, and multi-screen collaboration, inherently demand…
…Evaluating LLM Safety in Real-World Financial Scenarios (2026) SoK: Robustness in Large Language Models against Jailbreak Attacks (2026) MultiBreak: A Scalable and Diverse Multi-turn Jailbreak Benchmark for Evaluating LLM Safety…
…Agent Evaluation (2026) OpenComputer: Verifiable Software Worlds for Computer-Use Agents (2026) MobileGym: A Verifiable and Highly Parallel Simulation Platform for Mobile GUI Agent Research (2026) SimuWoB: Simulating Real-World Mobile Apps…
…A Real-World Benchmark and Physics-Guided Adaptation for Single-Image Relighting Published on May 12 Submitted by onurcan on May 13 Technical University of Denmark Authors: , Mehmet Onurcan Kaya , , Abstract WildRelight…
…A Unified Framework for Training, Evaluating, and Deploying GUI Agents (2026) SWE-Next: Scalable Real-World Software Engineering Tasks for Agents (2026) Claw-Eval: Toward Trustworthy Evaluation of Autonomous Agents (2026) Toward…
…Generated by Qwen/Qwen2.5-Coder-32B-Instruct Spatial reasoning is a fundamental capability for vision-language models (VLMs) deployed in real-world environments. However, visual observations are inherently limited representations of…
…Ultimately, we argue that AI safety audits must evolve to measure real-world leakage propensity, not just worst-case hacks, to give us a true, comprehensive picture of this phenomenon. This is…
…failure in the real world. Key highlights from our research: 🏭 Grounded in Reality: 2,049 items based on Chinese national standards (GB/T) and real industrial product records (evaluated across ZH, EN…
…Evaluating Agentic Multimodal Models via Compositional Visual Tool Chaining (2026) A3R: Agentic Affordance Reasoning via Cross-Dimensional Evidence in 3D Gaussian Scenes (2026) CresOWLve: Benchmarking Creative Problem-Solving Over Real-World Knowledge…
…Evaluating Robotic Manipulation Policies on Real-World Long-Horizon Tasks (2026) RoboFlow4D: A Lightweight Flow World Model Toward Real-Time Flow-Guided Robotic Manipulation (2026) Long-Term Memory for VLA-based Agents…