Salvato in:
Dettagli Bibliografici
Autori principali: Sia, Suzanna, Mueller, David, Duh, Kevin
Natura: Preprint
Pubblicazione: 2024
Soggetti:
Accesso online:https://arxiv.org/abs/2403.04510
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866917607603437568
author Sia, Suzanna
Mueller, David
Duh, Kevin
author_facet Sia, Suzanna
Mueller, David
Duh, Kevin
contents Self-supervised large language models have demonstrated the ability to perform Machine Translation (MT) via in-context learning, but little is known about where the model performs the task with respect to prompt instructions and demonstration examples. In this work, we attempt to characterize the region where large language models transition from in-context learners to translation models. Through a series of layer-wise context-masking experiments on \textsc{GPTNeo2.7B}, \textsc{Bloom3B}, \textsc{Llama7b} and \textsc{Llama7b-chat}, we demonstrate evidence of a "task recognition" point where the translation task is encoded into the input representations and attention to context is no longer necessary. We further observe correspondence between the low performance when masking out entire layers, and the task recognition layers. Taking advantage of this redundancy results in 45\% computational savings when prompting with 5 examples, and task recognition achieved at layer 14 / 32. Our layer-wise fine-tuning experiments indicate that the most effective layers for MT fine-tuning are the layers critical to task recognition.
format Preprint
id arxiv_https___arxiv_org_abs_2403_04510
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Where does In-context Translation Happen in Large Language Models
Sia, Suzanna
Mueller, David
Duh, Kevin
Computation and Language
Artificial Intelligence
Self-supervised large language models have demonstrated the ability to perform Machine Translation (MT) via in-context learning, but little is known about where the model performs the task with respect to prompt instructions and demonstration examples. In this work, we attempt to characterize the region where large language models transition from in-context learners to translation models. Through a series of layer-wise context-masking experiments on \textsc{GPTNeo2.7B}, \textsc{Bloom3B}, \textsc{Llama7b} and \textsc{Llama7b-chat}, we demonstrate evidence of a "task recognition" point where the translation task is encoded into the input representations and attention to context is no longer necessary. We further observe correspondence between the low performance when masking out entire layers, and the task recognition layers. Taking advantage of this redundancy results in 45\% computational savings when prompting with 5 examples, and task recognition achieved at layer 14 / 32. Our layer-wise fine-tuning experiments indicate that the most effective layers for MT fine-tuning are the layers critical to task recognition.
title Where does In-context Translation Happen in Large Language Models
topic Computation and Language
Artificial Intelligence
url https://arxiv.org/abs/2403.04510