The Curious Engineer

All blogs / AI / ML

Gaël Varoquaux

gael-varoquaux.info

Gaël Varoquaux, computer / data / health science

AI / MLPythonScience

’10’15’20’25
Posts
100 from 2 May 2010 to 9 Oct 2026
Posting
now and then (4 in the last year)
Kind
personal blog
Feed
https://gael-varoquaux.info/feeds/all.atom.xml

2026 4

  1. How much plagiarism is there in the mathematical genius of big AI?
  2. Will Intellectual Bulldozers Solve Math?
  3. Stepping up as probabl’s CSO to supercharge scikit-learn and its ecosystem
  4. 2025 highlights: AI research and code

2025 8

  1. Maïc, you lived 100 years, what changed?
  2. A national recognition; but science and open source are bitter victories
  3. TabICL: Pretraining the best tabular learner
  4. AI agents that use tools
  5. AIs that break down questions reason better
  6. Science must drive the narratives that shape society
  7. AI super-intelligent to play Go, and math?
  8. AI for health: the impossible necessity of unbiased data

2024 6

  1. 2024 highlights: of computer science and society
  2. When AIs must overcome the data
  3. Do AIs reason or recite?
  4. CARTE: toward table foundation models
  5. Skrub 0.2.0: tabular learning made easy
  6. Promoting open-source, from inria to :probabl.

2023 3

  1. People underestimate how impactful Scikit-learn continues to be
  2. Comité de l’intelligence artificielle: vision et stratégie nationale
  3. 2022, a new scientific adventure: machine learning for health and social sciences

2022 2

  1. My Mayavi story: discovering open source communities
  2. 2021 highlight: Decoding brain activity to new cognitive paradigms

2021 4

  1. 2021: my scientific year in review
  2. Hiring an engineer and post-doc to simplify data science on dirty data
  3. Hiring someone to develop scikit-learn community and industry partners
  4. 2020: my scientific year in review

2020 4

  1. Technical discussions are hard; a few tips
  2. Jean Dechoux, June 13rd 1923 – Feb 9th 2020
  3. Survey of machine-learning experimental methods at NeurIPS2019 and ICLR2020
  4. 2019: my scientific year in review

2019 3

  1. Comparing distributions: Kernels estimate good representations, l1 distances give good tests
  2. Getting a big scientific prize for open-source software
  3. 2018: my scientific year in review

2018 2

  1. A foundation for scikit-learn at Inria
  2. Sprint on scikit-learn, in Paris and Austin

2017 3

  1. Our research in 2017: personal scientific highlights
  2. Beyond computational reproducibility, let us aim for reusability
  3. Scikit-learn Paris sprint 2017

2016 4

  1. Our research in 2016: personal scientific highlights
  2. Data science instrumenting social media for advertising is responsible for todays politics
  3. Unison 2.48 binaries for ARM
  4. Better Python compressed persistence in joblib

2015 9

  1. Of software and Science. Reproducible science: what, why, and how
  2. Nilearn 0.2: more powerful machine learning for neuroimaging
  3. Job offer: data crunching brain functional connectivity for biomarkers
  4. MLOSS 2015: wising up to building open-source machine learning
  5. Nilearn sprint: hacking neuroimaging machine learning
  6. Software for reproducible science: let’s not have a misunderstanding
  7. MLOSS: machine learning open source software workshop @ ICML 2015
  8. Job offer: working on open source data processing in Python
  9. Euroscipy 2015: Call for paper

2014 8

  1. PRNI 2016: call for organization
  2. New website
  3. Improving your programming style in Python
  4. Hiring an engineer to mine large functional-connectivity databases
  5. Scikit-learn 2014 sprint: a report
  6. Scikit-learn 0.15 release: highlights
  7. Google summer of code projects for scikit-learn
  8. Hiring a programmer for a brain imaging machine-learning library

2013 2

  1. Publishing scientific software matters
  2. Scikit-learn 0.14 release: features and benchmarks

2012 8

  1. RIP John Hunter: the loss of a great man
  2. A journal promoting high-quality research code: dream and reality
  3. Update on scikit-learn: recent developments for machine learning in Python
  4. 3 Google summer of code for scikit-learn and more…
  5. The problems of low statistical power and publication bias
  6. Want features? Just code
  7. Book review: NumPy 1.5 Beginner’s guide
  8. Joblib beta release: fast compressed persistence + Python 3

2011 15

  1. Scikit-learn NIPS 2011 sprint: international thanks to our sponsors
  2. Cython example of exposing C-computed arrays in Python without data copies
  3. Python at scientific conferences
  4. Conference posters
  5. Hiring a junior developer on the scikit-learn
  6. My conference travels: Scipy 2011 and HBM 2011
  7. Euroscipy 2011: early bird deadline soon
  8. Hiring a junior engineer on the scikit-learn
  9. EuroScipy: the program is filling up, and the submission deadline nearing
  10. Scikit-learn sprint on April 1st
  11. Windows binaries for the scientific Python ecosystem
  12. Interested in parallel computing and statistics? We are looking for a post-doc
  13. EuroSciPy 2011: the dates are out - Aug 25-28, Paris
  14. Research jobs in France: the black humor of 2010 is the reality of 2011
  15. Scientific publication for software development

2010 15

  1. ICA versus PCA in the scikit-learn: the value of code over pictures
  2. Multitouch with VTK (and MedINRIA and Mayavi)
  3. Machine learning humour
  4. Scikit Learn coding sprint
  5. SVG Word map of countries
  6. Software design for maintainability
  7. Sprint Scikit learn in Paris
  8. Simple object signatures
  9. Euroscipy 2010: code, science, and a lot of fun
  10. Making posters for scientific conferences
  11. A simple LaTeX example
  12. Personal views on scientific computing
  13. EuroScipy abstract submission deadline extended
  14. EuroScipy is finally open for registration
  15. Status of the EuroScipy registration