Revisiting In-context Learning Inference Circuit in Large Language Models

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Cho, Hakaze, Kato, Mariko, Sakai, Yoshihiro, Inoue, Naoya
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915162271776768
author Cho, Hakaze
Kato, Mariko
Sakai, Yoshihiro
Inoue, Naoya
author_facet Cho, Hakaze
Kato, Mariko
Sakai, Yoshihiro
Inoue, Naoya
contents In-context Learning (ICL) is an emerging few-shot learning paradigm on Language Models (LMs) with inner mechanisms un-explored. There are already existing works describing the inner processing of ICL, while they struggle to capture all the inference phenomena in large language models. Therefore, this paper proposes a comprehensive circuit to model the inference dynamics and try to explain the observed phenomena of ICL. In detail, we divide ICL inference into 3 major operations: (1) Input Text Encode: LMs encode every input text (in the demonstrations and queries) into linear representation in the hidden states with sufficient information to solve ICL tasks. (2) Semantics Merge: LMs merge the encoded representations of demonstrations with their corresponding label tokens to produce joint representations of labels and demonstrations. (3) Feature Retrieval and Copy: LMs search the joint representations of demonstrations similar to the query representation on a task subspace, and copy the searched representations into the query. Then, language model heads capture these copied label representations to a certain extent and decode them into predicted labels. Through careful measurements, the proposed inference circuit successfully captures and unifies many fragmented phenomena observed during the ICL process, making it a comprehensive and practical explanation of the ICL inference process. Moreover, ablation analysis by disabling the proposed steps seriously damages the ICL performance, suggesting the proposed inference circuit is a dominating mechanism. Additionally, we confirm and list some bypass mechanisms that solve ICL tasks in parallel with the proposed circuit.
format Preprint
id arxiv_https___arxiv_org_abs_2410_04468
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Revisiting In-context Learning Inference Circuit in Large Language Models
Cho, Hakaze
Kato, Mariko
Sakai, Yoshihiro
Inoue, Naoya
Computation and Language
Artificial Intelligence
Machine Learning
In-context Learning (ICL) is an emerging few-shot learning paradigm on Language Models (LMs) with inner mechanisms un-explored. There are already existing works describing the inner processing of ICL, while they struggle to capture all the inference phenomena in large language models. Therefore, this paper proposes a comprehensive circuit to model the inference dynamics and try to explain the observed phenomena of ICL. In detail, we divide ICL inference into 3 major operations: (1) Input Text Encode: LMs encode every input text (in the demonstrations and queries) into linear representation in the hidden states with sufficient information to solve ICL tasks. (2) Semantics Merge: LMs merge the encoded representations of demonstrations with their corresponding label tokens to produce joint representations of labels and demonstrations. (3) Feature Retrieval and Copy: LMs search the joint representations of demonstrations similar to the query representation on a task subspace, and copy the searched representations into the query. Then, language model heads capture these copied label representations to a certain extent and decode them into predicted labels. Through careful measurements, the proposed inference circuit successfully captures and unifies many fragmented phenomena observed during the ICL process, making it a comprehensive and practical explanation of the ICL inference process. Moreover, ablation analysis by disabling the proposed steps seriously damages the ICL performance, suggesting the proposed inference circuit is a dominating mechanism. Additionally, we confirm and list some bypass mechanisms that solve ICL tasks in parallel with the proposed circuit.
title Revisiting In-context Learning Inference Circuit in Large Language Models
topic Computation and Language
Artificial Intelligence
Machine Learning
url https://arxiv.org/abs/2410.04468