Deep RLSP - Cheetah Forward

Learning to solve MuJoCo locomotion tasks from a single state: Cheetah running forward

Original (SAC)

Single state

GAIL (Baseline)

Deep RLSP

Ablation - AverageFeatures

10 states

GAIL (Baseline)

Deep RLSP

Ablation - AverageFeatures

Ablation - Waypoints

50 states

GAIL (Baseline)

Deep RLSP

Ablation - AverageFeatures

Ablation - Waypoints

Google Sites

Report abuse