Approximating How Single Head Attention Learns
Why do models often attend to salient words, and how does this evolve throughout training? We approximate model training as a two stage process: early on in training when the attention weights are uniform, the model learns to translate individual input word `i` to `o` if they co-occur frequently. La...
Saved in:
Main Authors | , , , |
---|---|
Format | Journal Article |
Language | English |
Published |
12.03.2021
|
Subjects | |
Online Access | Get full text |
Cover
Loading…
Summary: | Why do models often attend to salient words, and how does this evolve
throughout training? We approximate model training as a two stage process:
early on in training when the attention weights are uniform, the model learns
to translate individual input word `i` to `o` if they co-occur frequently.
Later, the model learns to attend to `i` while the correct output is $o$
because it knows `i` translates to `o`. To formalize, we define a model
property, Knowledge to Translate Individual Words (KTIW) (e.g. knowing that `i`
translates to `o`), and claim that it drives the learning of the attention.
This claim is supported by the fact that before the attention mechanism is
learned, KTIW can be learned from word co-occurrence statistics, but not the
other way around. Particularly, we can construct a training distribution that
makes KTIW hard to learn, the learning of the attention fails, and the model
cannot even learn the simple task of copying the input words to the output. Our
approximation explains why models sometimes attend to salient words, and
inspires a toy example where a multi-head attention model can overcome the
above hard training distribution by improving learning dynamics rather than
expressiveness. We end by discussing the limitation of our approximation
framework and suggest future directions. |
---|---|
DOI: | 10.48550/arxiv.2103.07601 |