Evaluation Design for Generative AI Agents in Social-Impact Domains
This presentation explores evaluation design for generative AI agents operating in social-impact domains, covering explainability, field fit, sustained operation, and accuracy.
Research Index
Research outputs are organized separately from books, media coverage, and public appearances. This page currently contains five placeholder research presentations and is structured for future DOI, PDF, slide, BibTeX, and related-project links.
Academic talks, workshops, posters, and technical reports.
This presentation explores evaluation design for generative AI agents operating in social-impact domains, covering explainability, field fit, sustained operation, and accuracy.
This talk discusses product design that combines decentralized identity, NFTs, and AI-assisted reflection to record, evaluate, and pass on regional contribution activities.
This technical report summarizes a RAG platform for research evidence collection, causal path building, and impact tracking from the perspective of retrieval quality, citability, and evaluation operations.
This presentation organizes how generative AI can support planning, feedback triage, and improvement cycles in civic participation products, with attention to operational design and risk management.
This poster examines information design for recording contributions, roles, and decision histories in DAO communities to support sustained participation and consensus building.