MAPL trains quadruped locomotion policies from LLM-generated multi-objective trajectory preferences and matches or exceeds expert-designed reward performance in four environments without manual reward engineering.
Primt: Preference-based reinforcement learning with multimodal feedback and trajectory synthesis from foundation mod- els,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.RO 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
MAPL: Multi-Objective Preference Learning for Robot Locomotion
MAPL trains quadruped locomotion policies from LLM-generated multi-objective trajectory preferences and matches or exceeds expert-designed reward performance in four environments without manual reward engineering.