Deep Q-Network ·
DQN
un
seel
.com · Pixels → Q(s,a) · Replay · Target net
Frames
0
Target sync
0
State
—
Online net Q(s,a;θ)
Target net θ⁻ (frozen)
Replay buffer
Chosen action (argmax)
▶ Play
←
→
🔇 Unmute
Reset
Un
seel
.com · Deep Q-Network