Implicit In-context Learning

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Li, Zhuowei, Xu, Zihao, Han, Ligong, Gao, Yunhe, Wen, Song, Liu, Di, Wang, Hao, Metaxas, Dimitris N.
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866929730574352384
author Li, Zhuowei
Xu, Zihao
Han, Ligong
Gao, Yunhe
Wen, Song
Liu, Di
Wang, Hao
Metaxas, Dimitris N.
author_facet Li, Zhuowei
Xu, Zihao
Han, Ligong
Gao, Yunhe
Wen, Song
Liu, Di
Wang, Hao
Metaxas, Dimitris N.
contents In-context Learning (ICL) empowers large language models (LLMs) to swiftly adapt to unseen tasks at inference-time by prefixing a few demonstration examples before queries. Despite its versatility, ICL incurs substantial computational and memory overheads compared to zero-shot learning and is sensitive to the selection and order of demonstration examples. In this work, we introduce Implicit In-context Learning (I2CL), an innovative paradigm that reduces the inference cost of ICL to that of zero-shot learning with minimal information loss. I2CL operates by first generating a condensed vector representation, namely a context vector, extracted from the demonstration examples. It then conducts an inference-time intervention through injecting a linear combination of the context vector and query activations back into the model's residual streams. Empirical evaluation on nine real-world tasks across three model architectures demonstrates that I2CL achieves few-shot level performance at zero-shot inference cost, and it exhibits robustness against variations in demonstration examples. Furthermore, I2CL facilitates a novel representation of task-ids, enhancing task similarity detection and fostering effective transfer learning. We also perform a comprehensive analysis and ablation study on I2CL, offering deeper insights into its internal mechanisms. Code is available at https://github.com/LzVv123456/I2CL.
format Preprint
id arxiv_https___arxiv_org_abs_2405_14660
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Implicit In-context Learning
Li, Zhuowei
Xu, Zihao
Han, Ligong
Gao, Yunhe
Wen, Song
Liu, Di
Wang, Hao
Metaxas, Dimitris N.
Machine Learning
Artificial Intelligence
Computation and Language
In-context Learning (ICL) empowers large language models (LLMs) to swiftly adapt to unseen tasks at inference-time by prefixing a few demonstration examples before queries. Despite its versatility, ICL incurs substantial computational and memory overheads compared to zero-shot learning and is sensitive to the selection and order of demonstration examples. In this work, we introduce Implicit In-context Learning (I2CL), an innovative paradigm that reduces the inference cost of ICL to that of zero-shot learning with minimal information loss. I2CL operates by first generating a condensed vector representation, namely a context vector, extracted from the demonstration examples. It then conducts an inference-time intervention through injecting a linear combination of the context vector and query activations back into the model's residual streams. Empirical evaluation on nine real-world tasks across three model architectures demonstrates that I2CL achieves few-shot level performance at zero-shot inference cost, and it exhibits robustness against variations in demonstration examples. Furthermore, I2CL facilitates a novel representation of task-ids, enhancing task similarity detection and fostering effective transfer learning. We also perform a comprehensive analysis and ablation study on I2CL, offering deeper insights into its internal mechanisms. Code is available at https://github.com/LzVv123456/I2CL.
title Implicit In-context Learning
topic Machine Learning
Artificial Intelligence
Computation and Language
url https://arxiv.org/abs/2405.14660