Building Open-Ended Embodied Agent via Language-Policy Bidirectional Adaptation.
Shaopeng ZhaiJie WangTianyi ZhangFuxian HuangQi ZhangMing ZhouJing HouYu LiuPublished in: CoRR (2024)
Keyphrases
- open ended
- intelligent agents
- action selection
- learning outcomes
- agent programming
- multi agent
- situated agents
- sensory motor
- multi agent systems
- agent model
- language learning
- agent receives
- external world
- multiagent systems
- autonomous agents
- pedagogical agents
- multiple agents
- markov decision process
- cognitive processes
- natural language
- multiple choice
- reward function
- state action
- programming language
- software agents
- learning environment
- reinforcement learning
- dynamic environments
- computer assisted language learning
- inquiry learning
- affect detection
- multimedia