Silver-Surfer-(GODOT)

Play the base game or view the three other models
Click any model to explore the live demo or repository.

Base Game

Original endless runner – HIGHLY RECCOMEND INSTALLING THE GAME FOR SMOOTHER AND BETTER GAMEPLAY

RL DQN Model

A real time Deep Q-Network reinforcement learning agent with prioritized experience replay, designed to interact with a game environment via socket communication(TCP)

GitHub →

Hybrid Agent

Behavior cloning + targeted exploration. Guaranteed improvement.

GitHub →

Rule‑Based Bot

Heuristic AI with lane discipline & speed adaptation. No learning.

GitHub →

Base Game

manual

Peak Score—

Avg Score—

Consistency—

📊 placeholder data – upload CSV later

RL DQN

deep Q

Peak Score—

Avg Score—

Consistency—

📊 placeholder – upload CSV later

Hybrid Agent

clone+explore

Peak Score—

Avg Score—

Consistency—

📊 placeholder – upload CSV later

Rule‑Based

heuristic

Peak Score—

Avg Score—

Consistency—

📊 placeholder – upload CSV later