Learn Before
Credit Assignment in AI Poetry Generation
An AI agent is tasked with writing a four-line poem. It generates the poem one word at a time. The agent only receives a score based on the final poem's creativity and coherence after all four lines are complete. Explain why this scoring system makes it difficult for the agent to determine the value of its individual word choices.
0
1
Tags
Ch.4 Alignment - Foundations of Large Language Models
Foundations of Large Language Models
Foundations of Large Language Models Course
Computing Sciences
Analysis in Bloom's Taxonomy
Cognitive Psychology
Psychology
Social Science
Empirical Science
Science
Related
Dense vs. Sparse Rewards
Reward Shaping as a Solution for Sparse Rewards
Transforming Sparse Rewards into Dense Supervision Signals
An AI is being trained to generate a multi-paragraph summary of a long document. The AI writes the summary one sentence at a time. A quality score is given only after the entire summary is complete. For each individual sentence generated before the final one, the score is zero. What is the most significant learning difficulty the AI will face due to this scoring method?
Training an Agent for a Text-Based Game
Credit Assignment in AI Poetry Generation
Methods for Mitigating Sparse Rewards