IJTCS 2021 | 分论坛日程：有意识的图灵机

编者按

第二届国际理论计算机联合大会（International Joint Conference on Theoretical Computer Science，IJTCS）将于2021年8月16日-20日在线上线下交互举行，由北京大学与中国工业与应用数学学会（CSIAM）、中国计算机学会（CCF）、国际计算机学会中国委员会（ACM China Council）联合主办，北京大学前沿计算研究中心承办，图灵奖获得者、中科院外籍院士、北京大学访问讲席教授John Hopcroft教授任大会主席。

本期带来“有意识的图灵机”分论坛的精彩介绍。

“有意识的图灵机”介绍

有意识的图灵机（Conscious Turing Machine，CTM）是由 Manuel Blum，Lenore Blum 和 Avrim Blum 提出的从理论计算机角度研究意识的产生的计算模型。CTM 基于图灵机这一理论计算机中的经典模型，对认知神经理论中的全局工作空间理论（Global Workspace Theory）进行尽可能简单的实现与建模。该模型可用来定义意识，解释意识的产生及与意识有关的一系列问题。有意识的图灵机可看成计算机科学家从计算机科学角度对“什么是意识？人类的意识如何产生”等有关人类自身发展问题的答案进行的一次探索。

本次会议将展示部分北京大学图灵班学生基于 CTM 这一计算模型，结合多方知识，从理论补充到实践应用进行的科研探索成果。

“有意识的图灵机”分论坛主席

 

Manuel Blum

Carnegie Mellon University

 

Lenore Blum

Carnegie Mellon University

 

Yurong Chen

Peking University

“有意识的图灵机”分论坛议程

时间：2021年8月18日

时间	讲者	报告题目
9:00- 9:55	Manuel Blum and Lenore Blum	Insights from the Conscious Turing Machine (CTM)
10:00- 10:25	丁   睿 郑凌骁	Mathematical Foundations of CTM
10:30- 10:55	吕   蓝 马   畅 苏博文	Doorway CTM: Explaining forgetting via Conscious Turing Machine
11:00- 11:25	李舒辰	MusiCTM: On Music Composing by CTM

“有意识的图灵机”分论坛报告简介

  

Insights from the Conscious Turing Machine (CTM)

Manuel Blum, 

Carnegie Mellon University

Lenore Blum, 

Carnegie Mellon University

Abstract

The quest to understand consciousness, once the purview of philosophers and theologians, is now actively pursued by scientists of many stripes.  In this talk, we discuss consciousness from the perspective of theoretical computer science (TCS), a branch of mathematics concerned with understanding the underlying principles of computation and complexity, especially the implications of resource limitations.  In the manner of TCS, we formalize the Global Workspace Theory (GWT) originated by cognitive neuroscientist Bernard Baars and further developed by him, Stanislas Dehaene, and others.  Our principal contribution lies in the precise formal definition of a Conscious Turing Machine (CTM).  We define the CTM in the spirit of Alan Turing’s simple yet powerful definition of a computer, the Turing Machine (TM).  We are not looking for a complex model of the brain nor of cognition but for a simple model of (the admittedly complex concept of) consciousness.

After defining CTM, we give a formal definition of consciousness in CTM.  We then suggest why the CTM has the feeling of consciousness.  The perspective given here provides a simple formal framework to employ tools from computational complexity theory and machine learning to further the understanding of consciousness.

This is joint work of Manuel, Lenore and Avrim Blum. 

Biography

Manuel Blum is a Professor of CS Emeritus at U.C. Berkeley (30 years) and CMU (20 years), and Visiting Chair Professor of CS at Peking University (2 years).  His recent work on Conscious Turing Machines satisfies his long standing desire to learn how to think, which he tried and continues trying to do by understanding brains, machines, mathematics, cognition and consciousness.  He is proudest of his talented family and the many exceptional students who got PhDs under his wing.

Lenore  Blum is an American computer scientist and mathematician, Founder and Professor Emeritus of the Math and Computer Science Department at Mills College (20 years),  Distinguished Career Professor of Computer Science at Carnegie Mellon University (20 years), and Visiting Chair Professor of CS at Peking University (2 years).  She is known for stunning mathematical discoveries in the Model Theory of Differential Fields, contributions to abstract models ofcomputation, a book with Cucker, Shub, and Smale entitled Complexity and Real Computation, Springer-Verlag, (1998), and current work on a model of consciousness called the Conscious Turing Machine (CTM), among many other contributions.

In 2005, she received the Presidential Award for Excellence in Science, Mathematics, and Engineering Mentoring, given her by president George W. Bush, and in 2018, she received the Simmons University Distinguished Alumnae Lifetime Achievement Award.

  

Mathematical Foundations of CTM

Rui Ding, Peking University

Lingxiao Zheng, Peking University

Abstract

The purpose of this project is to provide mathematical foundations related to CTM in an attempt to better model the concept of consciousness. The mathematics introduced here are indispensable since they equip the hypothetical machine with reasonable properties that enable us to omit low-level details which are incurred when we formalize this important yet elusive concept. Although many aspects of consciousness can be implied from this mechanic description, it is worth pointing out that no mathematically precise definition of the feeling of consciousness has been known so far.

In our project, our main work is to explore the competition function of CTM, which decides the chunk that wins the local competition. More specifically, we are going to decide which competition function is permutation invariant and find a universal form of permutation invariant competition function. This is important since permutation invariant means the permutation of LTM processors does not affect the probability at which each node is to be in STM after competition, thus we don't need to consider the permutation of LTM processors when implementing CTM. Apart from that, we are also going to explore Sleeping Experts Algorithm and discover some interesting mathematical properties of  it.

Biography

Rui Ding, majoring in CS, is currently a junior at PKU, whose research interests focus on network measurement and stochastic system within heavy-traffic regime. He is obsessed with works that are either sheerly theoretical or engineering, and is hoping to incorporate both beautifully in his future research.

Lingxiao Zheng is an undergraduate student in school of EECS, Peking University. He is currently advised by Prof. Tiejun Huang, and his research interest is visual information processing and neuromorphic computing.

   

Doorway CTM: Explaining forgetting via Conscious Turing Machine

Lan Lyu, Peking University

Chang Ma, Peking University

Bowen Su, Peking University

Abstract

"The Doorway Effect" refers to sudden forgetting that human exhibits when switching tasks. This phenomenon is key to understanding forgetting and has puzzled pyschologists for a long time. In this talk, we will share an explanation for the doorway effect by modeling the event boundary via Conscious Turing Machine.

Biography

Lan Lyu is a junior at Peking University. She had participated in some projects about indoor scene tasks, including Generative Automatic Furniture Arrangement via Dynamic Graph Learning, and agents rearrange objects in simulated indoor environment like AI2-thor. Currently, she is interested in human-computer interaction in real indoor scenes.

Chang Ma is currently a junior at Peking University, majored in Artificial Intelligence. Currently, she focuses on multiple machine learning topics, including computational biology, adversarial robustness, natural language processing, etc.

Bowen Su is currently a freshman at Peking University, majored in computer science. He is broadly interested in Algorithms and Machine Learning. Before college, he received a silver medal from the Chinese National Olympiad in Informatics (NOI) 2019.

 

MusiCTM: On Music Composing by CTM

Shuchen Li, Peking University

Abstract

Creativity is a powerful advantage for human beings. In this talk, we make an attempt to discover its interaction with consciousness by focusing on a specific aspect of creativity, music composing. Music composing was described as a three-phase process by Aranosian (Imagination, Cognition and Personality, 1981), that is, raw idea generation, formalization and development. We formulate these phases in the language of CTM (Conscious Turing Machine) and point out that consciousness and unconsciousness are both important in composing. In the first phase, flow state and associative thinking will help to produce raw auditory fragment unconsciously. In the second and the third phases, short-term memory and long-term memory interact, helping to formalize and learn the fragment, then determine how the next piece of music should come, in which consciousness involves. Finally, we hope that our work could inspire machine composing and extend to other forms of creation.

Biography

Shuchen Li is currently a junior student from CS Turing class, Peking University. His interests are in theoretical computer science, particularly Boolean function analysis and computational complexity.

关于IJTCS

回顾 → 对话邓小铁：在首届IJTCS中，我看到了中国计算理论的成长

日程 → 分论坛：区块链技术

日程 → 分论坛：多智能体强化学习

日程 → 分论坛：量子计算

日程 → 分论坛：算法与复杂性

日程 → 分论坛：人工智能与形式化方法

日程 → 分论坛：机器学习理论

日程 → 特   色：本科生论坛

IJTCS注册信息

大会现已正式面向公众开放注册！

观看线上报告：免费

通过在线观看直播的方式参与大会，可通过直播平台提问。

线上会议注册：

（普通）$100 /￥700

（学生）$50 /￥350*

获得所有Zoom会议参会链接，作为参会人在线参加全部会议，直接在线提问讨论并参与特设互动环节。

线下会议注册：

（普通）$200 / ￥1400

（学生）$100 / ￥700*

作为参会人在线下（北京大学）参加会议，与知名学者们面对面交流；同时享受线上注册的所有权益。

*因防疫要求，仅开放10个校外线下参会名额。

点击 ↓↓↓二维码↓↓↓ 跳转注册页面

*学生注册：网站上注册后需将学生证含有个人信息和学校信息的页拍照发送至[e-mail]，邮件主题格式为＂Student Registration+姓名+线上/线下"。

大会主席 

 

John Hopcroft

图灵奖获得者

中国科学院外籍院士

北京大学访问讲席教授

大会联合主席 

 

邓小铁

北京大学讲席教授

欧洲科学院外籍院士

ACM/IEEE Fellow

顾问委员会主席 

高  文

中国工程院院士

北京大学教授

梅  宏

中国科学院院士

CCF理事长

张平文

中国科学院院士

CSIAM理事长

北京大学教授

程序委员会主席 

孙晓明

中科院计算所

研究员

邓小铁

北京大学

讲席教授

李闽溟

香港城市大学

副教授

陆品燕

上海财经大学

教授

李  建

清华大学

副教授

组织单位 

合作媒体

大会赞助

联系人

大会赞助、合作等事宜

请联系

[e-mail]

010-62761029

大会网站

https://econcs.pku.edu.cn/ijtcs2021/index.htm

↑↑扫码直达大会官网↑↑

文字 | 吴宇森

—   版权声明  —

本微信公众号所有内容，由北京大学前沿计算研究中心微信自身创作、收集的文字、图片和音视频资料，版权属北京大学前沿计算研究中心微信所有；从公开渠道收集、整理及授权转载的文字、图片和音视频资料，版权属原作者。本公众号内容原作者如不愿意在本号刊登内容，请及时通知本号，予以删除。

点“阅读原文”转大会注册页面