UMAP update: new 'recursive' init scales to big data, reproducible multi-core runs
leland_mcinnes · x · 2026-09-30
UMAP author Leland McInnes detailed recent library improvements:
- New optimizers: fix a random seed while still using all cores for reproducible results.
- New "recursive" initialization: scales particularly well to large datasets that used to stall with spectral init.
- Hard negative mining via the negativesamplerange parameter.
- Bug fixes allow pushing repulsionstrength much higher.
A practical update for anyone visualizing large embeddings.
Related event: UMAP Gains Reproducible Optimizer and Recursive Initialization(2 posts)→
More from Research
- ReScraper: a 0.6B model replaces heuristic data-cleaning stacks, boosting pretraining up to 4.7% — XiongChenyan · 2026-09-30
- Microsoft Research finds LLMs show Dunning-Kruger-style overconfidence in coding — burkov · 2026-09-30
- PixAI launches anime foundation model Tsubaki.3, open-sources Tagger 1.0 with tech report — Level-Ninja-2492 · 2026-09-30
- Lean creator Leonardo de Moura on AI proofs: the Collatz exploit shows verified checkmarks can lie — Machine Learning Street Talk · 2026-09-30
- LLM-42 Paper at SOSP 2026 Brings Deterministic LLM Inference via Verified Speculation — tianyin_xu · 2026-09-30
- Frontier AI Is a Set, Not a Point: Jagged Capabilities May Be the Steady State — vsikka · 2026-09-30