Kristoffer Stensbo-Smidt

Postdoc, @CogSys, Technical University of Denmark. Working on AI that has a sense of uncertainty. Previously @CambridgeMLG.

[Kristoffer Stensbo-Smidt @kstensbo]

@VRLalchand Congratulations, Vidhi! 🎓 Very much looking forward to reading your thesis!

[Kristoffer Stensbo-Smidt @kstensbo]

@ikwess @sigmabayesian @jesfrellsen @WouterBoomsmaDK It was great having you, and thanks again for the fantastic talk! Incredibly clear and understandable introduction to neural processes!

[Wessel @ikwess]

Friday I spoke about autoregressive conditional neural processes at the Center for Basic Machine Learning in Life Science #MLLS, which resulted in some super interesting discussion.

Presentation:
https://t.co/drnc2Z7RGY

Paper:
https://t.co/uW11gcmP4R

[Sebastian Ober @sebastian_ober]

Late announcement, but I'm excited to announce that my PhD thesis is publicly available! In it I give my thoughts on how to approach variational inference in Bayesian neural nets, deep GPs, including discussion on the marginal likelihood and symmetries.

https://t.co/hZsAndVHeM

[Jes Frellsen @jesfrellsen]

Check out the cool work on Implicit VI scaled to large BNNs by @sigmabayesian, @kstensbo, @WouterBoomsmaDK and museum on Wednesday at #NeurIPS2023

[billy bubba bingus-drumpfus @wgrathwohl]

These homiez rule. Go talk to them.

[Jakub Tomczak @jmtomczak]

I am excited to announce that my book, "Deep Generative Modeling", is available online and in print (@SpringerNature): https://t.co/PkeTCJ0IuN
Code used in the book is freely available online: https://t.co/A7Peb2aERB (1/4)

[Sebastian Ober @sebastian_ober]

Excited to announce that my work with @laurence_ai on variational posteriors for deep Wishart processes has been accepted to #NeurIPS2021!

Paper: https://t.co/GTHNoAJ8r3

Come talk to us at location C3 of Poster Session 1 (tomorrow)! (1/n)

[Cambridge MLG @CambridgeMLG]

In tomorrow's CBL Alumni talk, we're happy to host our former PhD student @markvanderwilk, now assistant professor at @imperialcollege @ICComputing.

More details at
https://t.co/4TdzqLyUwr

Note: We just require attendees to have valid Zoom accounts (no registration required).

[Cambridge MLG @CambridgeMLG]

We're happy to present the CBL Alumni talk series!

🕑4-5pm BST every 2nd Friday

This week with @andrewgwils:
https://t.co/9CFya6Dkui

Next up: @markvanderwilk @shakir_za @DavidDuvenaud @FinaleDoshi @fhuszar

If you'd like to attend, please e-mail
[email protected]

[Kristoffer Stensbo-Smidt @kstensbo]

Interested in Bayesian inference and machine learning? Then you should set this new blog by @CambridgeMLG as your browser's start page!

[Wessel @ikwess]

@CambridgeMLG is launching a blog, featuring a first two-part post about what keeps a Bayesian awake at night by Richard E. Turner and me. 🧵

https://t.co/Gs8MhzAp6L

[Wessel @ikwess]

Annoyed by migrating from TF to PyTorch or the other way around? Perhaps now is the time for Jax? Then check out https://t.co/3qlEoSSBmB! LAB is a generic interface for linear algebra backends: code it once, run it on any backend. Just `pip install backends` :)

[Jose Miguel Hernández-Lobato @jmhernandez233]

PhD positions in Advanced Machine Learning at Cambridge
Application deadline: noon December 3, 2020.
Details about the application process can be found here:
https://t.co/2SwKfm9V8k

[Colby Itkowitz @ColbyItkowitz]

No one answered the phone at Four Seasons Total Landscaping. But oh man RIP their online reviews.

[Kristoffer Stensbo-Smidt @kstensbo]

This is a great idea! Also, taking some time to digest a paper before scoring it might give you a more nuanced perspective.

[Soheil Feizi @FeiziSoheil]

A suggestion to improve post-rebuttal reviewer engagements: don't ask reviewers to score (accept/reject) submissions *before* the rebuttal period, let them do this *after* discussions. By doing so, reviewers are more likely to engage with authors' responses & remain open minded.

[Arno Solin @arnosolin]

My #ICML2020 tutorial videos on "Machine Learning with Signal Processing" are now freely available:
I: https://t.co/FmXlqrlZC5
II: https://t.co/Gxr1dZLzSb
III: https://t.co/NaWrZD3ZFY
IV: https://t.co/wJMUcpKsVJ
Slides: https://t.co/kTUp0oZDpo

[Raphaël Millière @raphaelmilliere]

I asked GPT-3 to write a response to the philosophical essays written about it by @DrZimmermann, @rinireg @ShannonVallor, @add_hawk, @AmandaAskell, @dioscuri, David Chalmers, Carlos Montemayor, and Justin Khoo published yesterday by @DailyNousEditor. It's quite remarkable!

[Arno Solin @arnosolin]

Thank you all who participated in "Machine Learning with Signal Processing" during the morning run. The second run of the #ICML2020 tutorial will start soon.
Slides for the colour coded parts are now on available on https://t.co/kTUp0oZDpo

[Paper Digest @paperdigestorg]

Browse all #ICML2020 papers in 2 hours! Here is a one sentence highlight for every #ICML2020 Paper, plus related papers, patents. #MachineLearning #AI #computervision
https://t.co/mBODoZgMen

[Kristoffer Stensbo-Smidt @kstensbo]

@tyrell_turing @_hylandSL Not sure I agree with that. Because of the infinite number of functions, GPs become smoothers and are therefore not as flexible as NNs. I consider them different models; you trade flexibility for good generalisation, inference from little data, and well-calibrated uncertainties.