imarobot.org.ipynb — efficient MoE inference Python 3 · kernel idle
In [1]:
from research import Researcher imarobot_org = Researcher( name="imarobot.org", focus="memory-efficient MoE inference", status="independent researcher · open to collaborations", ) imarobot_org.introduce()
Out [1]:
imarobot.org

My research explores efficient neural network architectures and federated learning, with the goal of enabling a decentralised, open-source AI ecosystem. My work is shaped by a multidisciplinary background in AI research, embodied-AI simulation, and hands-on mechatronic design.

In [2]:
# most-cited first imarobot_org.papers.sort_values("citations").head(1)
Out [2]:
yeartitlevenue
2024Effectiveness of Large-Language Models in Recognizing Spatially Intensive Statistical Data
Designed and evaluated a human-in-the-loop automation for geospatial data reasoning.
SIT
In [3]:
for repo in imarobot_org.open_source: print(repo)
Out [3]:
Blend RLDesigned a lightweight open-source robotics simulator that allows researchers to use Blender scenes as physics-rich 3D environments to train AI-driven robot.★ 1
In [4]:
imarobot_org.contact()
Out [4]:
github homelab
In [5]: