📚 Education
Open tutors for low-resource languages
Train small, offline tutoring models for languages big labs ignore, so students learn in their mother tongue.
Impact drivers
density 74.0 / 100Each driver is normalised 0–1, then weighted. Nothing is hidden.
- Reach × 0.3580M people
- Urgency × 0.2550%
- Tractability × 0.2570%
- Evidence quality × 0.1560%
Diminishing returns
Total impact vs GPU-hours. The first hours buy the most; the dashed red line marks where returns start to saturate (~18k hrs).
Provable inputs
The exact parameters behind this score, fingerprinted so anyone can confirm the ranking used these numbers — not others.
project-params.sha256
$ sha256sum project-params.json5e2cd9fd2c6aae6b6dd3a82abb7bc3762e5494854248e49841302b4e2e0e020dfingerprint 5e2cd9fd…0e020d