CLOct 24, 2023

In-Context Learning Creates Task Vectors

DeepMind
arXiv:2310.15916v1301 citationsh-index: 32
Originality Incremental advance
AI Analysis

This provides insight into a fundamental AI learning paradigm, though it is incremental in explaining existing mechanisms.

The paper tackles the problem of understanding the mechanism behind in-context learning in large language models by showing that it compresses training data into a single task vector to modulate the transformer, supported by experiments across models and tasks.

In-context learning (ICL) in Large Language Models (LLMs) has emerged as a powerful new learning paradigm. However, its underlying mechanism is still not well understood. In particular, it is challenging to map it to the "standard" machine learning framework, where one uses a training set $S$ to find a best-fitting function $f(x)$ in some hypothesis class. Here we make progress on this problem by showing that the functions learned by ICL often have a very simple structure: they correspond to the transformer LLM whose only inputs are the query $x$ and a single "task vector" calculated from the training set. Thus, ICL can be seen as compressing $S$ into a single task vector $\boldsymbolθ(S)$ and then using this task vector to modulate the transformer to produce the output. We support the above claim via comprehensive experiments across a range of models and tasks.

Code Implementations1 repo
Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes