lmcinnes

4 ms·

1.
by lmcinnes 9mo ago
> alignment tooling is fascinating, as we increasingly want to re-fit->embed over time as our envs change and compare, eg, day-over-day analysis. This area is not well-defined yet common for anyone operational so seems ripe for innova
2.
Visualizing a topic model of NeurIPS 2025 Papers(newsletter.languagemodels.co)
2 points by lmcinnes 11mo ago|0 comments
3.
SoarXiv: Soar through ArXiv paper space(soarxiv.org)
3 points by lmcinnes 1y ago|1 comments
4.
by lmcinnes 1y ago
> And algorithms can only predict content that you've seen before. It'll never surprise you with something different. It keeps you in a little bubble. This is not true at all, algorithms can predict things you haven't se
5.
by lmcinnes 2y ago
Assuming you have a dimension-reduction or manifold learning tool of choice (UMAP,PacMAP,t-SNE,PyMDE,etc.) then DataMapPlot ( https://datamapplot.readthedocs.io/en/latest/ ) is a library specifically designed to mak
6.
by lmcinnes 2y ago
If you just want in-memory then PyNNDescent ( https://github.com/lmcinnes/pynndescent ) can work pretty well. It should install easily with pip, works well at the scales you mention, and supports a large number of metric
7.
Show HN: Exploring five million Hacker News posts(lmcinnes.github.io)
3 points by lmcinnes 2y ago|0 comments
8.
Show HN: An interactive explorable map of 5M Hacker News stories(lmcinnes.github.io)
3 points by lmcinnes 2y ago|0 comments
9.
The Landscape of Machine Learning on ArXiv(lmcinnes.github.io)
31 points by lmcinnes 2y ago|10 comments
10.
Show HN: An Interactive Map of Machine Learning Papers on ArXiv(lmcinnes.github.io)
1 points by lmcinnes 3y ago|0 comments
11.
Show HN: DataMapPlot for visualizing large corpora of documents(github.com)
3 points by lmcinnes 3y ago|0 comments
12.
Show HN: DataMapPlot for beautiful plots of data maps(datamapplot.readthedocs.io)
2 points by lmcinnes 3y ago|0 comments
13.
by lmcinnes 3y ago
For suitable specialized cases thins can be quite efficient. For persistent H_0 of VR-complexes in low-dimensional space there is an O(N log(N)) algorithm for N data points; that's decently fast. If you want H_1 I believe (but cannot p
14.
by lmcinnes 5y ago
It is really not that much slower for training (see the paper), and if you are interested in pipelines the difference is not so great considering you are looking at a one off training time vs. lots of inference.
15.
by lmcinnes 5y ago
If this is a thing you want to be able to do efficiently then ParametricUMAP (see [docs]( https://umap-learn.readthedocs.io/en/latest/parametric_umap.... ) and [the paper]( https://arxiv.org/abs/
16.
by lmcinnes 6y ago
Density based clustering with high dimensional data will tend to struggle. This is because, in high enough dimensions, you need a lot of samples to see any density. Also distances start to look very similar (from the curse of dimensionali
17.
An Approx Nearest Neighbor Algorithm: How PyNNDescent Works(pynndescent.readthedocs.io)
2 points by lmcinnes 6y ago|0 comments
18.
Approx Nearest Neighbor Search: How PyNNDescent Works(pynndescent.readthedocs.io)
2 points by lmcinnes 6y ago|0 comments
19.
Comparing Deep Neural Networks with UMAP Tour(tiga1231.github.io)
1 points by lmcinnes 6y ago|0 comments
20.
by lmcinnes 7y ago
I suspect that this is because GPT-2 doesn't have any overarching narrative that it is piecing together. Ultimately it is like a super-powerful Markov based text generator -- predicting what comes next from what has come before. It has
21.
Understanding Hdbscan and Density-Based Clustering(towardsdatascience.com)
3 points by lmcinnes 7y ago|0 comments
22.
Embed All the Things! Embedding Beers, Breweries, and Beer Styles (among others)(youtube.com)
2 points by lmcinnes 7y ago|0 comments
23.
Understanding the UMAP dimension reduction algorithm(pair-code.github.io)
1 points by lmcinnes 7y ago|0 comments
24.
Understanding the UMAP dimension reduction algorithm(pair-code.github.io)
2 points by lmcinnes 7y ago|0 comments
25.
Topological methods for unsupervised learning problems [video](slideslive.com)
57 points by lmcinnes 8y ago|11 comments
26.
A Bluffer's Guide to Dimension Reduction(youtube.com)
2 points by lmcinnes 8y ago|0 comments
27.
Analysing Audio with Wavenet, MFCCs, UMAP, T-SNE and PCA(medium.com)
2 points by lmcinnes 9y ago|0 comments
28.
Show HN: Mapping and Analysing Subreddits(lmcinnes.github.io)
21 points by lmcinnes 9y ago|0 comments
29.
Mapping Subreddits with T-SNE Variant LargeVis(lmcinnes.github.io)
3 points by lmcinnes 9y ago|0 comments
30.
Mapping and Analysing SubReddits Using Python(lmcinnes.github.io)
2 points by lmcinnes 9y ago|0 comments

More ›