归档

2026  2

February  1

SafeVerse Open Source: Building a Safe and Trustworthy Embodied AI ‘Twin Training Ground’

February 10, 2026 · 6 min · Shanghai Artificial Intelligence Laboratory · Safety & Trustworthy Center

January  1

Loop as a Bridge: Can Looped Transformers Truly Link Representation Space and Natural Language Outputs?

January 16, 2026 · 4 min · Guanxu Chen, Dongrui Liu, Jing Shao

2025  6

July  6

SafeWork-F1: A Frontier AI Risk Management Framework

July 21, 2025 · 11 min · Center for Safe&Trustworthy AI

SafeWork-T1: A Safety Reasoning Training Accelerator for Multimodal Large Models

July 21, 2025 · 3 min · Center for Safe&Trustworthy AI

The intelligent agent lurking in your phone

July 18, 2025 · 7 min · Center for Safe&Trustworthy AI

SafeWork-E1: The Untiring Second Brain - Safe & Trustworthy Foundation Model for Clean Energy

July 14, 2025 · 5 min · Center for Safe&Trustworthy AI

SafeWork-R1: Coevolving Safety and Intelligence under the AI-45° Law

July 12, 2025 · 7 min · Center for Safe&Trustworthy AI

SafeWork-V1: Towards Formally Verifiable AI

July 12, 2025 · 3 min · Center for Safe&Trustworthy AI