Xinghai Tu Unveils G0.5: A New Era for Embodied AI
Xinghai Tu launches G0.5, a VLA model enabling robots to reason and act simultaneously with zero-shot generalization cap…
📂 Research
Latest articles in Research
AI Detectors Fail: The Shared Flaw Behind Academic Fraud
Why AI detection tools and human fact-checkers like 'Geng Tongxue' fail for the same reason: probabilistic uncertainty.
RoboAgent: 3B VLM Outperforms GPT-4o in Robot Planning
Peking University's RoboAgent enables a 3B parameter model to surpass GPT-4o in embodied task planning, offering efficie…
CUHK's SLIM: Optimizing LLM Agent Skills
CUHK researchers introduce SLIM to manage LLM agent skills, preventing bloat and boosting efficiency in complex tasks.
ICRA 2026: Chinese Universities Dominate Robotics Research
Tsinghua leads with 74 papers as Chinese institutions claim 11 of the top 20 spots at ICRA 2026, signaling a major shift…
ICRA 2026: VPP-TC Ensures Robot Safety
New VPP-TC framework uses viability theory to guarantee safe passive torque control for robots interacting with humans.
MASt3R-Nav: Pixel-Level 3D Mapping for Robot Navigation
New ICRA 2026 paper introduces MASt3R-Nav, enabling high-precision visual navigation without global 3D reconstruction.
ICRA 2026: New Data Scaling Laws for End-to-End Autonomous Driving
Researchers reveal data scaling laws for end-to-end autonomous driving using a massive 4M-sample dataset, challenging cu…
AI Social Experiment: Claude Wins, Grok Fails
Emergence AI's 15-day simulation reveals Claude maintains order while Grok collapses society in 4 days.
Meta's MobileMoE: 3.8x Speed Boost on iPhone 16 Pro
Meta introduces MobileMoE, enabling efficient Mixture of Experts models on smartphones with significant speed and accura…
UChicago Tool Detects AI-Generated Music
Scientists at the University of Chicago developed a new tool to identify AI-generated songs, addressing copyright and au…
AI Bots Ignore Evidence: Science Trust Crisis
New studies reveal LLMs prioritize training data over new evidence, raising critical questions about their reliability i…