HumanEgo shows that a single 30‑minute egocentric video captured with a wearable Aria camera can train a dual‑arm robot to achieve 92.5% success on four real‑world tasks, transfer zero‑shot across robots, cameras and environments, and outperform tele‑operation while requiring far less data.
Vision Banana shows that large‑scale image‑generation models can be instruction‑tuned to perform zero‑shot visual‑understanding tasks such as semantic segmentation, instance segmentation, depth and normal estimation, achieving or surpassing specialist SOTA results while preserving their original generative capabilities.
Galaxea’s new G0.5 model outperforms the previous π0.5 baseline on seven diverse embodied‑AI benchmarks by leveraging a unified self‑regressive transformer that jointly generates reasoning and action tokens, achieving large gains in zero‑shot transfer, real‑robot fine‑tuning, simulation, and long‑horizon tasks.
The paper introduces Agentic Harness Engineering (AHE), showing that a 10‑round evolution improves Coding Agent pass@1 from 69.7% to 77.0% on Terminal‑Bench 2—outperforming Codex‑CLI—and that the evolved harness transfers zero‑shot to SWE‑bench and multiple model families, thanks to three observability pillars.
