Transformer 度=23
MemGPT 度=15
Generative Agents 度=15
ReAct 度=13
LLM-based Agent Architecture 度=12
Memory Module 度=9
Planning Module 度=8
Scaled Dot-Product Attention 度=4
Multi-Head Attention 度=3
Sinusoidal Positional Encoding 度=2
Position-wise FFN 度=2
Profiling Module 度=2
Action Module 度=2
Memory Stream 度=2
Retrieval Function 度=2
Reflection Mechanism 度=2
Agent Planning 度=2
Chain-of-Thought Prompting 度=2
RNN Encoder-Decoder 度=1
ConvS2S 度=1
Chain-of-Thought (CoT) 度=1
Tree of Thoughts (ToT) 度=1
LLM+P 度=1
Reflexion 度=1
Prompt Engineering 度=1
Agent Fine-tuning 度=1
Mechanism Engineering 度=1
Virtual Context Management 度=1
Self-directed Memory Management 度=1
Agent Reacting 度=1
Thought-Action-Observation Loop 度=1
Self-Consistency (CoT-SC) 度=1
ReAct + CoT-SC Hybrid 度=1
Few-Shot In-Context Learning 度=1
Inner Monologue 度=1
ReAct Bootstrapped Finetuning 度=1