Transformers as Statisticians: Provable In-Context Learning with In-Context Algorithm Selection.
Yu BaiFan ChenHuan WangCaiming XiongSong MeiPublished in: CoRR (2023)
Keyphrases
- learning algorithm
- dynamic programming
- learning speed
- matching algorithm
- detection algorithm
- optimization algorithm
- times faster
- learning tasks
- cost function
- active learning
- selection algorithm
- recognition algorithm
- context aware
- preprocessing
- learning phase
- improved algorithm
- experimental evaluation
- incremental learning
- computational complexity
- learning systems
- learning rules
- reinforcement learning
- contextual information
- prior knowledge
- classification algorithm
- theoretical analysis
- convergence rate
- particle swarm optimization
- input data
- computational cost
- significant improvement
- optimal solution
- objective function