Rabbinic Embedding Benchmark
Evaluate Hebrew/Aramaic embedding models for cross-lingual retrieval
At Sefaria, we utilize AI to broaden access to Jewish texts. Interests: Hebrew (of all periods), Rabbinic Aramaic, named entity recognition, citation extraction, recommendation algorithms, multilingual embedding, and RAG to support generative models with factual citations. We're open source and as open-data as we can be - always looking to support development of digital Torah ventures.
Nonprofit / Open Data / Jewish Educational Technology
Sefaria is dedicated to building the future of Jewish learning in an open and participatory way. We are creating a free living library of Jewish texts and their interconnections - Bible, Talmud, Midrash, Chasidut, Kabbalah, Liturgy, Responsa... Our goal is to support learners, educators, and developers by providing open access to our digital texts and tools that enable new forms of engagement with this ancient tradition.
Utilize our datasets, contribute translations, or help develop our open-source platform.