About Us

Strategic Game Deep Learning Research Lab

The DeepLab team comprises researchers in deep learning, game theory, and reinforcement learning, focused on imperfect-information games and multi-agent decision-making applied to strategic games.

Why Strategic Games?

The world's most popular 'strategy games' aren't Chess or Go — they're strategic games. Tens of millions of players, rich strategy ecosystems, complete research literature, and the deep-learning problems here are harder than Go:

  • ① Imperfect information (can't see opponents' hands)
  • ② N-player games (cooperation × competition)
  • ③ Real-time decisions (no time for deep search)

Three Research Threads

RESEARCH 01
Imperfect-Information Game

Strategic reasoning under hidden information

RESEARCH 02
Multi-Agent Decision

Strategy learning for N-player games

RESEARCH 03
Online Continual Learning

Models that absorb new mechanics post-deployment

01

Deep Learning Lab · Not General-Purpose

We don't compete with ChatGPT or Claude on general intelligence. DeepLab is a deep-learning lab for strategic games — models 100% optimized for strategic play.

02

Platform Approach · Not a Single Tool

Industry solver tools mostly focus on a single game. DeepLab follows a "single product form, full variant matrix" strategy — covering multiple mechanic families.

03

Research-Driven · Beyond Product

We publish research notes, paper deep-dives, and academic collaboration opportunities on deeplab.pro/#news.