Zihan Li
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Research Interests
Zihan Li has a total of 4 co-authors. Here are the top co-authors by number of shared publications:
- April Yi Wang
ETH Zurich, Zurich, Switzerland
1 shared publication - Kristin Williams
Computer Science Department, Emory University, Atlanta, USA
1 shared publication - Lahari Goswami
ETH Zurich, Zurich, Switzerland
1 shared publication - Yuqing Feng
TU Munich, Munich, Germany
1 shared publication
Latest Publications
- Open Access
- 13 April 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-6
"Rules or Data? Gotta Catch 'Em All!": A Tangible Game for Youth AI Literacy
Adolescents often learn about AI through single-model demonstrations, such as training a classifier, which can blur how rule-based and data-driven systems differ and why those differences matter for errors, bias, and trust. We present Pokémon AI Adventure, a tangible, narrative-driven learning game in which students build and compare two robot partners: RuleBot, which follows feature-based rules, and DataBot, which learns from examples. To complete zone challenges and capture Pokémon, students assemble rule sets by selecting and ordering physical Clue Cards, and curate training datasets using Package Cards while managing noise, imbalance, and missing Pokémon types. Through these activities, the game supports learning about how rules and data shape model behavior, how accuracy and bias arise from human choices, and how to develop more reasoned trust in different AI systems.
- 13 April 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-6
Open Access"Rules or Data? Gotta Catch 'Em All!": A Tangible Game for Youth AI Literacy
Adolescents often learn about AI through single-model demonstrations, such as training a classifier, which can blur how rule-based and data-driven systems differ and why those differences matter for errors, bias, and trust. We present Pokémon AI Adventure, a tangible, narrative-driven learning game in which students build and compare two robot partners: RuleBot, which follows feature-based rules, and DataBot, which learns from examples. To complete zone challenges and capture Pokémon, students assemble rule sets by selecting and ordering physical Clue Cards, and curate training datasets using Package Cards while managing noise, imbalance, and missing Pokémon types. Through these activities, the game supports learning about how rules and data shape model behavior, how accuracy and bias arise from human choices, and how to develop more reasoned trust in different AI systems.