Triple Phase Transitions: Understanding the Learning Dynamics of Large Language Models from a Neuroscience Perspective

TL;DR

This paper investigates the abrupt phase transitions in large language models during training, drawing parallels with neuroscience to understand their learning dynamics and emergent behaviors.

Contribution

It introduces a novel interpretation of LLM learning dynamics, identifying three key phase transitions and their relation to brain-like alignment and task performance.

Findings

Three phase transitions identified during training

Alignment with the brain surges, then diverges, then re-aligns

Insights into the mechanisms underlying emergent behaviors

Abstract

Large language models (LLMs) often exhibit abrupt emergent behavior, whereby new abilities arise at certain points during their training. This phenomenon, commonly referred to as a ''phase transition'', remains poorly understood. In this study, we conduct an integrative analysis of such phase transitions by examining three interconnected perspectives: the similarity between LLMs and the human brain, the internal states of LLMs, and downstream task performance. We propose a novel interpretation for the learning dynamics of LLMs that vary in both training data and architecture, revealing that three phase transitions commonly emerge across these models during training: (1) alignment with the entire brain surges as LLMs begin adhering to task instructions Brain Alignment and Instruction Following, (2) unexpectedly, LLMs diverge from the brain during a period in which downstream task accuracy temporarily stagnates Brain Detachment and Stagnation, and (3) alignment with the brain reoccurs as LLMs become capable of solving the downstream tasks Brain Realignment and Consolidation. These findings illuminate the underlying mechanisms of phase transitions in LLMs, while opening new avenues for interdisciplinary research bridging AI and neuroscience.