reinforcement learning
/ˌriːɪnˈfɔːrsmənt ˌlɜːrnɪŋ/
Learning through trial and error with rewards and penalties
reinforcement learning in a sentence
“Reinforcement learning enabled the AI to master complex games.”
Origin of reinforcement learning
From Latin re- (again) + in- (in) + fortis (strong), meaning to strengthen again
Related Words
transformer
A neural network architecture using self-attention mechanisms
embedding
A dense vector representation of data in continuous space
fine-tuning
Adapting a pre-trained model for a specific task
inference
Using a trained model to make predictions on new data
hallucination
When an AI generates false or fabricated information
prompt engineering
Crafting inputs to elicit desired outputs from AI models