Experimental Anki algorithm uses RWKV to predict retrievability, claims to crush rivals
panickssery · x · 2026-11-20
A user shared an experimental Anki spaced-repetition algorithm that uses an RWKV model (hybrid RNN/transformer) to directly predict retrievability, claiming it outperforms every existing algorithm by a wide margin. It can be run via an addon or a custom Anki fork maintained by the author. The original tweet was quoted to ask Grok whether it's real.
More from Apps
- How to find AI ad creatives that actually sell before paid teams copy them — eptwts · 2026-11-20
- Berkeley Haas pilots Rubrica AI to grade handwritten exams, cutting time 40-60% — _akpiper · 2026-11-20
- Google rolls out collapsible sidebar on desktop to revisit recent AI searches — rmstein · 2026-11-20
- LavaFlow bug-fixing feature enters early access, bundled with all paid plans of Volcano — testingcatalog · 2026-11-20
- Garry Tan shares GBrain cases: AI working overnight, 850 pages in 3 weeks — garrytan · 2026-11-20
- Granola switches default model to GPT-6 Sol, criticized as dumber and 2x pricier than GPT-6.1 — gabriel1 · 2026-11-20