Eight Things to Know about Large Language Models

TL;DR

This paper reviews eight key insights about large language models, highlighting their capabilities, unpredictability, interpretability challenges, and the limitations of human performance comparisons.

Contribution

It provides a comprehensive survey of eight critical and often surprising points about LLMs, emphasizing the need for careful consideration of their capabilities and limitations.

Findings

LLMs become more capable with increased investment.

Many behaviors emerge unpredictably as investment grows.

Interactions with LLMs can be misleading.

Abstract

The widespread public deployment of large language models (LLMs) in recent months has prompted a wave of new attention and engagement from advocates, policymakers, and scholars from many fields. This attention is a timely response to the many urgent questions that this technology raises, but it can sometimes miss important considerations. This paper surveys the evidence for eight potentially surprising such points: 1. LLMs predictably get more capable with increasing investment, even without targeted innovation. 2. Many important LLM behaviors emerge unpredictably as a byproduct of increasing investment. 3. LLMs often appear to learn and use representations of the outside world. 4. There are no reliable techniques for steering the behavior of LLMs. 5. Experts are not yet able to interpret the inner workings of LLMs. 6. Human performance on a task isn't an upper bound on LLM performance. 7. LLMs need not express the values of their creators nor the values encoded in web text. 8. Brief interactions with LLMs are often misleading.