Learn Before
Two-Step Post-Pre-training Alignment Process
A common practice in developing Large Language Models involves implementing two distinct alignment stages after the initial pre-training on extensive unlabeled datasets.
0
1
Tags
Ch.2 Generative Models - Foundations of Large Language Models
Foundations of Large Language Models
Foundations of Large Language Models Course
Computing Sciences
Related
Two-Step Post-Pre-training Alignment Process
A technology company claims it can create a perfectly helpful and harmless AI assistant by simply pre-training a model on an exhaustive dataset containing all books, articles, and websites ever published. They argue that such a comprehensive dataset would make any subsequent training phase to align the model's behavior unnecessary. Which of the following statements provides the most critical evaluation of this claim's primary flaw?
Rationale for Post-Training Alignment
Critique of a Pre-training-Only Approach
Learn After
Justifying the LLM Development Pipeline
A team of AI researchers is developing a new large language model intended for public use. Arrange the following high-level stages of their development process into the correct chronological order, starting from the initial training.
An AI development team has just completed the initial, large-scale training of a new language model on a massive dataset of text and code from the internet. What is the primary distinction in purpose between this initial training phase and the alignment stages that typically follow?