Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Contextual Position Encoding: Learning to Count What's Important

About

The attention mechanism is a critical component of Large Language Models (LLMs) that allows tokens in a sequence to interact with each other, but is order-invariant. Incorporating position encoding (PE) makes it possible to address by position, such as attending to the i-th token. However, current PE methods use token counts to derive position, and thus cannot generalize to higher levels of abstraction, such as attending to the i-th sentence. In this paper, we propose a new position encoding method, Contextual Position Encoding (CoPE), that allows positions to be conditioned on context by incrementing position only on certain tokens determined by the model. This allows more general position addressing such as attending to the $i$-th particular word, noun, or sentence. We show that CoPE can solve the selective copy, counting and Flip-Flop tasks where popular position embeddings fail, and improves perplexity on language modeling and coding tasks.

Olga Golovneva, Tianlu Wang, Jason Weston, Sainbayar Sukhbaatar• 2024

Related benchmarks

TaskDatasetResultRank
Grid Navigation2D Grid Navigation S 256/128/0.8 (OOD-sparse)
Accuracy70
12
Grid Navigation2D Grid Navigation Sequence length 128, grid width 64 (IID)
Accuracy76
12
Grid Navigation2D Grid Navigation D 64/32/0.2 (OOD-dense)
Accuracy86
12
Grid Navigation1D Grid Navigation Sequence length 128, grid width 64 (IID)
Accuracy88
12
Grid Navigation1D Grid Navigation 64/32/0.2 (OOD-dense D)
Accuracy91
12
Grid Navigation1D Grid Navigation OOD-sparse S 256/128/0.8
Accuracy65
12
Selective-Copy TaskSelective-Copy OOD Dense split: 64/128 blank/non-blank
Accuracy100
8
Selective-Copy TaskSelective-Copy OOD Sparse 256 128 blank non-blank
Accuracy100
8
Selective-Copy TaskSelective-Copy IID 128/128 blank/non-blank
Accuracy100
8
Grid Navigation5D Navigation (IID)
Accuracy94
5
Showing 10 of 15 rows

Other info

Follow for update