Japan's Humanitext wins $816,667 Schmidt Sciences HAVI grant to build verified classical-text knowledge graph
SebastianNehrd2 · x · 2026-10-15
Humanitext, a project led by Iwata Naoya of Nagoya University, is one of 19 winners (out of 501 applications from 42 countries) in Schmidt Sciences' Humanities and AI Virtual Institute (HAVI) 2026 program, receiving $816,667 over three years.
- The project builds a verifiable knowledge graph of relationships (citation, allusion, paraphrase, critique) among 2,300+ Greek and Latin classics, stored in Wikibase + Neo4j with provenance.
- Researchers will first hand-build 500 relationships as a gold standard, then have LLMs propose candidate links for expert scholars to accept, fix, or reject one by one.
- Goal: 10,000+ verified relationships in three years, plus a GraphRAG vs RAG comparison on citation accuracy with external classicists.
- All code, models, and benchmarks will be open-sourced.
More from AGI Musings
- Game studios can sell privately hosted realistic games to AI labs as training environments — dejavucoder · 2026-10-15
- Retellings Use New Perspectives to Replace Romance with Therapeutic Diagnosis — eigenrobot · 2026-10-15
- CHI Reviewer: AI Slop Is Real, but Unusually Rigorous Papers Still Getting Through — IanArawjo · 2026-10-15
- ASML Sells Zero Lithography Machines into Europe as US Hosts 75%+ of AI Compute — NinaDSchick · 2026-10-15
- Into the Unknown: CMU benchmark finds 114 of 135 AI-written research papers fail integrity checks — arankomatsuzaki · 2026-10-15
- LLMs-as-Media Erode Producer Power; Author Ties It to Securities Reform — curious_vii · 2026-10-15