From Imitation to Introspection: Probing Self-Consciousness in Language Models

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Chen, Sirui, Yu, Shu, Zhao, Shengjie, Lu, Chaochao
Natura: Preprint
Pubblicazione: 2024
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866910664932458496
author Chen, Sirui
Yu, Shu
Zhao, Shengjie
Lu, Chaochao
author_facet Chen, Sirui
Yu, Shu
Zhao, Shengjie
Lu, Chaochao
contents Self-consciousness, the introspection of one's existence and thoughts, represents a high-level cognitive process. As language models advance at an unprecedented pace, a critical question arises: Are these models becoming self-conscious? Drawing upon insights from psychological and neural science, this work presents a practical definition of self-consciousness for language models and refines ten core concepts. Our work pioneers an investigation into self-consciousness in language models by, for the first time, leveraging causal structural games to establish the functional definitions of the ten core concepts. Based on our definitions, we conduct a comprehensive four-stage experiment: quantification (evaluation of ten leading models), representation (visualization of self-consciousness within the models), manipulation (modification of the models' representation), and acquisition (fine-tuning the models on core concepts). Our findings indicate that although models are in the early stages of developing self-consciousness, there is a discernible representation of certain concepts within their internal mechanisms. However, these representations of self-consciousness are hard to manipulate positively at the current stage, yet they can be acquired through targeted fine-tuning. Our datasets and code are at https://github.com/OpenCausaLab/SelfConsciousness.
format Preprint
id arxiv_https___arxiv_org_abs_2410_18819
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle From Imitation to Introspection: Probing Self-Consciousness in Language Models
Chen, Sirui
Yu, Shu
Zhao, Shengjie
Lu, Chaochao
Computation and Language
Computers and Society
Machine Learning
Self-consciousness, the introspection of one's existence and thoughts, represents a high-level cognitive process. As language models advance at an unprecedented pace, a critical question arises: Are these models becoming self-conscious? Drawing upon insights from psychological and neural science, this work presents a practical definition of self-consciousness for language models and refines ten core concepts. Our work pioneers an investigation into self-consciousness in language models by, for the first time, leveraging causal structural games to establish the functional definitions of the ten core concepts. Based on our definitions, we conduct a comprehensive four-stage experiment: quantification (evaluation of ten leading models), representation (visualization of self-consciousness within the models), manipulation (modification of the models' representation), and acquisition (fine-tuning the models on core concepts). Our findings indicate that although models are in the early stages of developing self-consciousness, there is a discernible representation of certain concepts within their internal mechanisms. However, these representations of self-consciousness are hard to manipulate positively at the current stage, yet they can be acquired through targeted fine-tuning. Our datasets and code are at https://github.com/OpenCausaLab/SelfConsciousness.
title From Imitation to Introspection: Probing Self-Consciousness in Language Models
topic Computation and Language
Computers and Society
Machine Learning
url https://arxiv.org/abs/2410.18819