corpus.blog
56,966 blogs · [ { "id": "01a08724-bd03-73a9-927b-894a8e575b95", "title": "AI for humans", "url": "https://alex.smola.org/posts/49-ai-for-humans/", "published_at": "2026-08-09T15:00:00+00:00" }, { "id": "01a08724-bd04-7296-8b61-3ac316268303", "title": "Losing the thread", "url": "https://alex.smola.org/posts/48-losing-the-thread/", "published_at": "2026-07-05T15:00:00+00:00" }, { "id": "01a08724-bd04-7296-8b61-3ac316f49f5a", "title": "The autopsy: porting MXNet to Blackwell and Apple Silicon", "url": "https://alex.smola.org/posts/47-mxnet-autopsy/", "published_at": "2026-06-18T15:00:00+00:00" }, { "id": "01a08724-bd04-7296-8b61-3ac31749a799", "title": "Raising MXNet from the Attic", "url": "https://alex.smola.org/posts/46-mxnet-zombie/", "published_at": "2026-06-17T15:00:00+00:00" }, { "id": "01a08724-bd04-7296-8b61-3ac3182799ff", "title": "Efficiency in LLMs", "url": "https://alex.smola.org/posts/45-mlss-efficiency/", "published_at": "2026-06-12T15:00:00+00:00" }, { "id": "01a08724-bd04-7296-8b61-3ac319117efc", "title": "Fast projections require a PHD", "url": "https://alex.smola.org/posts/43-fjlt/", "published_at": "2026-06-09T15:00:00+00:00" }, { "id": "01a08724-bd04-7296-8b61-3ac319a10943", "title": "Random projections, three ways", "url": "https://alex.smola.org/posts/42-random-projections/", "published_at": "2026-06-08T15:00:00+00:00" }, { "id": "01a08724-bd04-7296-8b61-3ac31a319a46", "title": "Digital humans", "url": "https://alex.smola.org/posts/41-digital-humans/", "published_at": "2026-06-06T15:00:00+00:00" }, { "id": "01a08724-bd04-7296-8b61-3ac31a36973e", "title": "The effective sample size", "url": "https://alex.smola.org/posts/40-effective-sample-size/", "published_at": "2026-06-01T15:00:00+00:00" }, { "id": "01a08724-bd04-7296-8b61-3ac31a715a80", "title": "ProactBench", "url": "https://alex.smola.org/posts/39-proactbench/", "published_at": "2026-05-31T15:00:00+00:00" }, { "id": "01a08724-bd04-7296-8b61-3ac31acab07f", "title": "What your assistant didn’t say", "url": "https://alex.smola.org/posts/38-proactivity/", "published_at": "2026-05-28T15:00:00+00:00" }, { "id": "01a08724-bd04-7296-8b61-3ac31b01014b", "title": "Beating \\(2^n - 1\\) for InfiniBand", "url": "https://alex.smola.org/posts/37-infiniband-2n/", "published_at": "2026-05-26T15:00:00+00:00" }, { "id": "01a08724-bd04-7296-8b61-3ac31b61b4e2", "title": "Fair representations are pancakes", "url": "https://alex.smola.org/posts/36-pancake-theorem/", "published_at": "2026-05-25T15:00:00+00:00" }, { "id": "01a08724-bd04-7296-8b61-3ac31c091733", "title": "You don’t need all the benchmarks", "url": "https://alex.smola.org/posts/34-benchmark-selection/", "published_at": "2026-05-24T15:00:00+00:00" }, { "id": "01a08724-bd04-7296-8b61-3ac31cec5073", "title": "The Pokémon Theorem", "url": "https://alex.smola.org/posts/35-pokemon-theorem/", "published_at": "2026-05-24T15:00:00+00:00" }, { "id": "01a08724-bd04-7296-8b61-3ac31d6a2326", "title": "Dive into Deep Learning", "url": "https://alex.smola.org/posts/33-d2l/", "published_at": "2019-01-14T15:00:00+00:00" }, { "id": "01a08724-bd04-7296-8b61-3ac31df2bd26", "title": "Leaving CMU", "url": "https://alex.smola.org/posts/32-leaving-cmu/", "published_at": "2016-05-31T15:00:00+00:00" }, { "id": "01a08724-bd04-7296-8b61-3ac31e8c32e3", "title": "Distributing Data in a Parameter Server", "url": "https://alex.smola.org/posts/31-distributing-data/", "published_at": "2013-01-14T15:00:00+00:00" }, { "id": "01a08724-bd04-7296-8b61-3ac31f6e056f", "title": "100 Terabytes, 5 Billion Documents, 10 Billion Parameters, 1 Billion Inserts/s", "url": "https://alex.smola.org/posts/30-100-terabytes/", "published_at": "2012-12-14T15:00:00+00:00" }, { "id": "01a08724-bd04-7296-8b61-3ac31fa29948", "title": "Beware the bandwidth gap - speeding up optimization", "url": "https://alex.smola.org/posts/28-bandwidth-gap/", "published_at": "2012-10-14T15:00:00+00:00" }, { "id": "01a08724-bd04-7296-8b61-3ac31fe38adc", "title": "The Weisfeiler-Leman algorithm and estimation on graphs", "url": "https://alex.smola.org/posts/27-weisfeiler-lehman/", "published_at": "2012-09-14T15:00:00+00:00" }, { "id": "01a08724-bd04-7296-8b61-3ac32068da88", "title": "In defense of keeping data private", "url": "https://alex.smola.org/posts/26-private-data/", "published_at": "2012-08-14T15:00:00+00:00" }, { "id": "01a08724-bd04-7296-8b61-3ac32135374a", "title": "MLSS Purdue", "url": "https://alex.smola.org/posts/25-mlss-purdue/", "published_at": "2012-07-14T15:00:00+00:00" }, { "id": "01a08724-bd04-7296-8b61-3ac3215249a8", "title": "Random numbers in constant storage", "url": "https://alex.smola.org/posts/24-random-numbers/", "published_at": "2012-06-14T15:00:00+00:00" }, { "id": "01a08724-bd04-7296-8b61-3ac321ff0073", "title": "The Neal Kernel and Random Kitchen Sinks", "url": "https://alex.smola.org/posts/22-neal-kernel/", "published_at": "2012-04-14T15:00:00+00:00" }, { "id": "01a08724-bd04-7296-8b61-3ac3225d28a0", "title": "Big Learning: Algorithms, Systems, and Tools for Learning at Scale", "url": "https://alex.smola.org/posts/21-big-learning/", "published_at": "2011-08-14T15:00:00+00:00" }, { "id": "01a08724-bd04-7296-8b61-3ac323444570", "title": "Introduction to Graphical Models", "url": "https://alex.smola.org/posts/20-graphical-models/", "published_at": "2011-08-09T15:00:00+00:00" }, { "id": "01a08724-bd04-7296-8b61-3ac32349cb7a", "title": "Distributed synchronization with the distributed star", "url": "https://alex.smola.org/posts/19-distributed-star/", "published_at": "2011-08-04T15:00:00+00:00" }, { "id": "01a08724-bd04-7296-8b61-3ac323f1ae31", "title": "Speeding up Latent Dirichlet Allocation", "url": "https://alex.smola.org/posts/18-fast-lda/", "published_at": "2011-07-31T15:00:00+00:00" }, { "id": "01a08724-bd04-7296-8b61-3ac3242f7e20", "title": "Bloom Filters", "url": "https://alex.smola.org/posts/17-bloom-filter/", "published_at": "2011-07-29T15:00:00+00:00" } ] posts Claim your blog
Back to alex.smola.org
Blog · corpus.blog/blogs/alex.smola.org/posts

alex.smola.org

alex.smola.org

2026

AI for humansoriginal ↗
9 Aug 2026
Losing the threadoriginal ↗
5 Jul 2026
The autopsy: porting MXNet to Blackwell and Apple Siliconoriginal ↗
18 Jun 2026
Raising MXNet from the Atticoriginal ↗
17 Jun 2026
Efficiency in LLMsoriginal ↗
12 Jun 2026
Fast projections require a PHDoriginal ↗
9 Jun 2026
Random projections, three waysoriginal ↗
8 Jun 2026
Digital humansoriginal ↗
6 Jun 2026
The effective sample sizeoriginal ↗
1 Jun 2026
ProactBenchoriginal ↗
31 May 2026
What your assistant didn’t sayoriginal ↗
28 May 2026
Beating \(2^n - 1\) for InfiniBandoriginal ↗
26 May 2026
Fair representations are pancakesoriginal ↗
25 May 2026
You don’t need all the benchmarksoriginal ↗
24 May 2026
The Pokémon Theoremoriginal ↗
24 May 2026

2019

Dive into Deep Learningoriginal ↗
14 Jan 2019

2016

Leaving CMUoriginal ↗
31 May 2016

2013

Distributing Data in a Parameter Serveroriginal ↗
14 Jan 2013

2012

100 Terabytes, 5 Billion Documents, 10 Billion Parameters, 1 Billion Inserts/soriginal ↗
14 Dec 2012
Beware the bandwidth gap - speeding up optimizationoriginal ↗
14 Oct 2012
The Weisfeiler-Leman algorithm and estimation on graphsoriginal ↗
14 Sept 2012
In defense of keeping data privateoriginal ↗
14 Aug 2012
MLSS Purdueoriginal ↗
14 Jul 2012
Random numbers in constant storageoriginal ↗
14 Jun 2012
The Neal Kernel and Random Kitchen Sinksoriginal ↗
14 Apr 2012

2011

Big Learning: Algorithms, Systems, and Tools for Learning at Scaleoriginal ↗
14 Aug 2011
Introduction to Graphical Modelsoriginal ↗
9 Aug 2011
Distributed synchronization with the distributed staroriginal ↗
4 Aug 2011
Speeding up Latent Dirichlet Allocationoriginal ↗
31 Jul 2011
Bloom Filtersoriginal ↗
29 Jul 2011