我们研究人工智能
最底层的技术。
LOOPDOT IS A RESEARCH TEAM WORKING AT THE FOUNDATIONAL LAYER OF ARTIFICIAL INTELLIGENCE — THE ARCHITECTURES, SYSTEMS, AND MECHANICS BENEATH EVERY APPLICATION.
在循环迭代中，打磨基础。
LoopDot 相信真正的智能突破，源于对模型架构、训练系统与推理机制的深层理解，而非表层应用的堆砌。我们专注底层，让每一层技术都经得起推敲。
WE BELIEVE REAL BREAKTHROUGHS COME FROM A DEEP UNDERSTANDING OF ARCHITECTURE, TRAINING SYSTEMS, AND INFERENCE MECHANICS — NOT FROM STACKING APPLICATIONS ON THE SURFACE. WE REFINE THE FUNDAMENTALS, LOOP BY LOOP.
四个核心方向
模型架构Model Architecture
探索神经网络结构设计的边界，从注意力机制到新型拓扑结构。
EXPLORING THE BOUNDARIES OF NETWORK DESIGN.
大规模训练系统Training Systems
研究分布式训练、显存优化与训练效率背后的系统工程。
SYSTEMS ENGINEERING BEHIND LARGE-SCALE TRAINING.
推理与效率优化Inference & Efficiency
让模型在真实世界的算力与延迟约束下，跑得更快、更省、更稳。
FASTER, LEANER, MORE RELIABLE UNDER CONSTRAINTS.
学习算法与理论Learning & Theory
从优化理论到泛化边界，追问一个问题：为什么它有效。
ASKING WHY IT WORKS.
论文与技术报告
Adaptive Adversarial Intelligence
A Unified Mathematical Theory of Learning Competition in Military Artificial Intelligence
1 Institute of Defence and Strategic Studies (IDSS), S. Rajaratnam School of International Studies (RSIS), Nanyang Technological University (NTU), Singapore
2 LoopDot AI Research
ABSTRACT Artificial intelligence is transforming military decision making from static policy optimization into continual learning against adaptive intelligent opponents.