MAE WIRE · Ars Technica · AI

DeepMind reveals Gemini Robotics 2.5 and open safety benchmark "ASIMOV-jailbreak"

来源 Ars Technica · AI 作者 Ryan Whitwam 2026-11-10 · 8 天前 English
DeepMind reveals Gemini Robotics 2.5 and open safety benchmark "ASIMOV-jailbreak"
Summary · 摘要

The new model is up to twice as successful at some tasks.

阅读全文 arstechnica.com