서브메뉴
검색
Youth in the Loop: Harnessing Children's Exploration, Causal Reasoning, and Data to Shape the Development and Evaluation of Artificial Intelligence
Youth in the Loop: Harnessing Children's Exploration, Causal Reasoning, and Data to Shape the Development and Evaluation of Artificial Intelligence
상세정보
- 자료유형
- 학위논문 서양
- 최종처리일시
- 20260202103547
- ISBN
- 9798288866562
- DDC
- 136
- 저자명
- Kosoy, Eliza.
- 서명/저자
- Youth in the Loop: Harnessing Childrens Exploration, Causal Reasoning, and Data to Shape the Development and Evaluation of Artificial Intelligence
- 발행사항
- [Sl] : University of California, Berkeley, 2025
- 발행사항
- Ann Arbor : ProQuest Dissertations & Theses, 2025
- 형태사항
- 78 p
- 주기사항
- Source: Dissertations Abstracts International, Volume: 87-01, Section: B.
- 주기사항
- Advisor: Gopnik, Alison.
- 학위논문주기
- Thesis (Ph.D.)--University of California, Berkeley, 2025.
- 초록/해제
- 요약As artificial intelligence becomes increasingly woven into daily life, understanding how youth perceive, interact with, and inspire these technologies is critical. This dissertation brings children's perspectives into the heart of AI research, arguing for a "Youth-in-the-Loop" approach. Across four studies, we explore how children's innate curiosity and flexible cognition provide unique insights for both improving AI models and benchmarking their development. We compare children's exploratory behavior with reinforcement learning agents, revealing profound differences in strategies and generalization. We investigate how children learn causal structures and show that they outperform current computational models in flexible hypothesis generation. We adapt classical developmental psychology experiments to evaluate large language models, proposing novel metrics for assessing alignment with human cognition. Finally, we conduct empirical studies capturing children's direct interactions with generative AI, highlighting their optimism and creative engagement. Together, these works underscore the importance of integrating children's data and developmental perspectives in building, training, and evaluating AI systems for a future that serves and empowers the next generation.
- 일반주제명
- Developmental psychology
- 일반주제명
- Experimental psychology
- 일반주제명
- Behavioral psychology
- 키워드
- Causal learning
- 키워드
- Generative AI
- 기타저자
- University of California, Berkeley Psychology
- 기본자료저록
- Dissertations Abstracts International. 87-01B.
- 전자적 위치 및 접속
- 로그인 후 원문을 볼 수 있습니다.
MARC
008260126s2025 us c eng d■001000017357690
■00520260202103547
■006m o d
■007cr#unu||||||||
■020 ▼a9798288866562
■035 ▼a(MiAaPQ)AAI32041378
■040 ▼aMiAaPQ▼cMiAaPQ
■0820 ▼a136
■1001 ▼aKosoy, Eliza.
■24510▼aYouth in the Loop: Harnessing Children's Exploration, Causal Reasoning, and Data to Shape the Development and Evaluation of Artificial Intelligence
■260 ▼a[Sl]▼bUniversity of California, Berkeley▼c2025
■260 1▼aAnn Arbor▼bProQuest Dissertations & Theses▼c2025
■300 ▼a78 p
■500 ▼aSource: Dissertations Abstracts International, Volume: 87-01, Section: B.
■500 ▼aAdvisor: Gopnik, Alison.
■5021 ▼aThesis (Ph.D.)--University of California, Berkeley, 2025.
■520 ▼aAs artificial intelligence becomes increasingly woven into daily life, understanding how youth perceive, interact with, and inspire these technologies is critical. This dissertation brings children's perspectives into the heart of AI research, arguing for a "Youth-in-the-Loop" approach. Across four studies, we explore how children's innate curiosity and flexible cognition provide unique insights for both improving AI models and benchmarking their development. We compare children's exploratory behavior with reinforcement learning agents, revealing profound differences in strategies and generalization. We investigate how children learn causal structures and show that they outperform current computational models in flexible hypothesis generation. We adapt classical developmental psychology experiments to evaluate large language models, proposing novel metrics for assessing alignment with human cognition. Finally, we conduct empirical studies capturing children's direct interactions with generative AI, highlighting their optimism and creative engagement. Together, these works underscore the importance of integrating children's data and developmental perspectives in building, training, and evaluating AI systems for a future that serves and empowers the next generation.
■590 ▼aSchool code: 0028.
■650 4▼aDevelopmental psychology
■650 4▼aExperimental psychology
■650 4▼aBehavioral psychology
■653 ▼aCausal learning
■653 ▼aChild-AI interaction
■653 ▼aDevelopmental benchmarking
■653 ▼aGenerative AI
■653 ▼aReinforcement Learning from Human Feedback
■653 ▼aYouth-in-the-Loop
■690 ▼a0620
■690 ▼a0800
■690 ▼a0623
■690 ▼a0384
■71020▼aUniversity of California, Berkeley▼bPsychology.
■7730 ▼tDissertations Abstracts International▼g87-01B.
■790 ▼a0028
■791 ▼aPh.D.
■792 ▼a2025
■793 ▼aEnglish
■85640▼uhttp://www.riss.kr/pdu/ddodLink.do?id=T17357690▼nKERIS▼z이 자료의 원문은 한국교육학술정보원에서 제공합니다.
