서브메뉴
검색
Systematicity and Explicit Representations in Humans and Machines
Systematicity and Explicit Representations in Humans and Machines
상세정보
- 자료유형
- 학위논문 서양
- 최종처리일시
- 20250211153100
- ISBN
- 9798346390428
- DDC
- 793.73
- 서명/저자
- Systematicity and Explicit Representations in Humans and Machines
- 발행사항
- [Sl] : Stanford University, 2024
- 발행사항
- Ann Arbor : ProQuest Dissertations & Theses, 2024
- 형태사항
- 227 p
- 주기사항
- Source: Dissertations Abstracts International, Volume: 86-05, Section: A.
- 주기사항
- Advisor: McClelland, Jay.
- 학위논문주기
- Thesis (Ph.D.)--Stanford University, 2024.
- 초록/해제
- 요약Recognizing systematic and structural symmetries is a fundamental aspect of human cognition, crucial for mastering language, mathematical reasoning, and problem-solving. This dissertation investigates the characteristics, contributing factors, and potential mechanisms that foster systematicity in both humans and machines. First, I present evidence that demonstrate a clear bimodality in humans' ability to learn and generalize during an abstract reasoning task, alongside differences in individual performance that covary with explicitness of task representations and prior mathematical education. Second, I build on these insights using transformer-based neural network models by integrating abstract relational planning into their training regimen, significantly boosting their efficacy in tasks involving graph and mathematical reasoning. Finally, I investigate the mechanisms underlying explicit abstract representations using a neural stochastic dynamical systems model that offers an algorithmic account of the sub-symbolic processes that implement symbol-like cognitive phenomena. Collectively, these studies offer a multi-faceted view into the behavioral characteristics and mechanisms underlying systematic reasoning in both human and artificial cognitive systems.
- 일반주제명
- Puzzles
- 일반주제명
- Dynamical systems
- 일반주제명
- Education
- 기타저자
- Stanford University.
- 기본자료저록
- Dissertations Abstracts International. 86-05A.
- 전자적 위치 및 접속
- 로그인 후 원문을 볼 수 있습니다.
MARC
008250123s2024 us c eng d■001000017164899
■00520250211153100
■006m o d
■007cr#unu||||||||
■020 ▼a9798346390428
■035 ▼a(MiAaPQ)AAI31652065
■035 ▼a(MiAaPQ)Stanfordvh602jq1834
■040 ▼aMiAaPQ▼cMiAaPQ
■0820 ▼a793.73
■1001 ▼aNam, Andrew Joohun.
■24510▼aSystematicity and Explicit Representations in Humans and Machines
■260 ▼a[Sl]▼bStanford University▼c2024
■260 1▼aAnn Arbor▼bProQuest Dissertations & Theses▼c2024
■300 ▼a227 p
■500 ▼aSource: Dissertations Abstracts International, Volume: 86-05, Section: A.
■500 ▼aAdvisor: McClelland, Jay.
■5021 ▼aThesis (Ph.D.)--Stanford University, 2024.
■520 ▼aRecognizing systematic and structural symmetries is a fundamental aspect of human cognition, crucial for mastering language, mathematical reasoning, and problem-solving. This dissertation investigates the characteristics, contributing factors, and potential mechanisms that foster systematicity in both humans and machines. First, I present evidence that demonstrate a clear bimodality in humans' ability to learn and generalize during an abstract reasoning task, alongside differences in individual performance that covary with explicitness of task representations and prior mathematical education. Second, I build on these insights using transformer-based neural network models by integrating abstract relational planning into their training regimen, significantly boosting their efficacy in tasks involving graph and mathematical reasoning. Finally, I investigate the mechanisms underlying explicit abstract representations using a neural stochastic dynamical systems model that offers an algorithmic account of the sub-symbolic processes that implement symbol-like cognitive phenomena. Collectively, these studies offer a multi-faceted view into the behavioral characteristics and mechanisms underlying systematic reasoning in both human and artificial cognitive systems.
■590 ▼aSchool code: 0212.
■650 4▼aPuzzles
■650 4▼aDynamical systems
■650 4▼aEducation
■690 ▼a0515
■71020▼aStanford University.
■7730 ▼tDissertations Abstracts International▼g86-05A.
■790 ▼a0212
■791 ▼aPh.D.
■792 ▼a2024
■793 ▼aEnglish
■85640▼uhttp://www.riss.kr/pdu/ddodLink.do?id=T17164899▼nKERIS▼z이 자료의 원문은 한국교육학술정보원에서 제공합니다.
