Yarin @yaringal Twitter profile

Last Seen Profiles

@Twitch_ANZ

@DJV_JVBB

@AngelaJone38425

@Mary63439656377

@cqrson

@Mahadevbhakt002

@IUCN_Plastics

@official4SYTETV

@mayorcool_segun

@zda_jon

@sebasRenaud17

@Zervas_

@alinalal_

@anne_clarke

@asuna_mizushima

@bangaritan

@abuelyazia

@aki_7stars

@jwerthein

@geyminae

@NaplesSafety

@BARRACKp

@ZHRANCO_SA

@_Brayonie

@fenwickfriars

@NoahPro28

@codenamekyle

@calxandr

@Effymom

@arxxxxmand

@Hektorestrada

@court_ossan_bot

@Kerimaix

@Daisybelll

@BoxingKSA

@hypohate

Yarin

@yaringal

5 years

Really excited to release Bayesian Deep Learning Benchmarks - please share with others who you think might like this, and have a look at the blog/repo/colab: This work was done over a period of a year and a half by many collaborators @OATML_Oxford

9

407

1K

Yarin

@yaringal

4 months

I'm hiring! I'm building 4 research groups under me at AISI (formerly the UK's Taskforce on Frontier AI) to work on foundational AI safety research. [1/5]

Introducing the AI Safety Institute

www.gov.uk

14

153

810

Yarin

@yaringal

5 years

All the slides from my Bayesian Deep Learning tutorial at MLSS 2019 Moscow, including a practical in Active Learning with jupyter notebooks (practical credit: Ivan Nazarov), are now online

4

126

584

Yarin

@yaringal

8 years

New blog post! "Uncertainty in Deep Learning" - also my PhD thesis and lots of new results

5

263

556

Yarin

@yaringal

4 years

I really like blog posts which try to teach the reader old ideas. This one is about ways to visualise concepts in information theory, mentioned (cautiously) in solution 8.8 in Mackay's book (remember, positive areas can be negative quantities!) By @BlackHC

Better intuition for information theory

Better visual explanations of information theoretic quantities like entropy and mutual information using I-diagrams. Based on Raymond W. Yeung's "A new outlook on Shannon's information measures"...

www.blackhc.net

1

120

540

Yarin

@yaringal

3 years

Did you know that you can beat Deep Ensemble Uncertainty with a single deterministic net? We prove that softmax nets can't normally capture epistemic uncertainty, but with an appropriate inductive bias any pre-trained net can implicitly capture uncertainty

6

78

480

Yarin

@yaringal

6 years

Dear research community, please keep machine learning open access @NipsConference @icmlconf @iclr2017 JMLR

Nature Machine Intelligence

Nature Machine Intelligence will publish high-quality original research and reviews in a wide range of topics in machine learning, robotics and AI. The ...

www.nature.com

3

124

366

Yarin

@yaringal

4 years

Very happy to be named one of MIT Technology Review @techreview Europe's 35 under 35. Many thanks to all my collaborators over the years!

23

7

307

Yarin

@yaringal

3 years

This year's #BDL workshop at @NeurIPSConf will focus on the reliability of BDL in downstream tasks, with invited talks from practitioners and the two NeurIPS BDL challenges Please consider submitting extended abstracts by Oct 1, or posters by Dec 1

1

73

298

Yarin

@yaringal

4 years

Our car below never saw roundabouts at training time. But using dropout ensembles' epistemic uncertainty we can choose the best worst-case plan to follow at deployment We put code online to make it as easy as MNIST to plug & play your own BDL tools:

Yarin

@yaringal

4 years

Can autonomous 🚘 identify, recover from, and adapt to distribution shifts? We play with BDL and robust control to get cars to recover&adapt when they don't know what to do At ICML with @filangelos @ptigas @rowantmc @nick_rhinehart @svlevine 📄🎞️🕸️💻:

3

38

173

4

44

294

Yarin

@yaringal

4 years

The Bayesian Deep Learning workshop website has been updated with accepted papers and schedule #BDL2019

4

48

251

Yarin

@yaringal

7 years

Bayesian neural networks uncertainty can be used to distinguish adversarial from non-adversarial images! new results

6

82

241

Yarin

@yaringal

6 years

Is there a "reproducible research" website? Which aggregates all open source reproductions of arXiv papers and people get credit for reproducing results? Might lead to proper incentive for others to spend time recreating exps. Useful for criticism, discussions, and follow up exps

16

55

241

Yarin

@yaringal

4 years

Bayesian deep learning workshop talks from NeuRIPS 2019 are already available online: #BDL2019 #NeurIPS2019

Yarin Gal

slideslive.com

0

71

238

Yarin

@yaringal

7 years

Our videos from the Bayesian Deep Learning Workshop at NIPS are now online! #BDL2016

Bayesian Deep Learning Workshop NIPS 2016

Videos for the NIPS 2016 Bayesian Deep Learning Workshop http://bayesiandeeplearning.org/

www.youtube.com

1

116

232

Yarin

@yaringal

7 years

Excited to announce that from this October I will be Associate Professor of machine learning at @UniofOxford @CompSciOxford department...

16

9

220

Yarin

@yaringal

3 years

Curious to see what a rejected #ICLR2021 paper with scores 8,6,6,4 looks like? (might as well get some PR) Presenting our work on tractable objectives for information bottlenecks! We propose IB bounds which scale to imagenet & are really easy to implement

Unpacking Information Bottlenecks: Surrogate Objectives for Deep...

The Information Bottleneck principle offers both a mechanism to explain how deep neural networks train and generalize, as well as a regularized objective with which to train models. However...

openreview.net

9

20

217

Yarin

@yaringal

6 years

Excited to announce our new research group at @CompSciOxford : Oxford Applied and Theoretical Machine Learning Group (OATML) Have a look at the website and ping me at NIPS if you'd like to join us!

6

35

217

Yarin

@yaringal

6 years

Awesome line-up for this year's Bayesian deep learning workshop @NipsConference , with this year's theme "deep learning uncertainty in real-world applications"

Bayesian Deep Learning Workshop | NeurIPS 2021

Bayesian Deep Learning Workshop at NeurIPS 2021 — Tuesday, December 14, 2021, Virtual.

bayesiandeeplearning.org

1

56

204

Yarin

@yaringal

7 years

How powerful are Graph Convolutional Networks? - An introduction to neural networks on graphs via @thomaskipf

0

70

199

Yarin

@yaringal

9 months

I'm testing llama 2: Bob and Alice went to the pub. Bob forgot his keys and went back to the car to take them. Alice waited for Bob for a long time and decided to go home at the end. When Bob got back and looked for Alice he couldn't find her. Where is Bob? Entertaining responses

36

16

200

Yarin

@yaringal

4 years

This year the #BDL workshop will take a new form, and will be organised as a @NeurIPSConf European event together with @ELLISforEurope We invite researchers to submit posters for presentation at the event (**deadline: December 1, 2020**)

1

42

195

Yarin

@yaringal

4 years

Yay we're the 6th most cited paper from ICML from the past 5 years. Many thanks to everyone using these tools and to all my awesome collaborators over the years!

4

5

193

Yarin

@yaringal

3 years

Big announcement: we classify ~72k protein variants, previously of unknown significance, using unsupervised ML trained on evolutionary and human sequences. We perform on par on variants studied in lab exps. Result of a year-long collaboration with @deboramarks lab @harvardmed

Jonathan Frazer

@Jonnygfrazer

3 years

1/n I'm excited to share our preprint *Large-scale clinical interpretation of genetic variants using evolutionary data and deep learning* from a great collab with @NotinPascal @MafaldaFigDias @AidanNGomez @kpgbrock @yaringal ( @OATML_Oxford ) @deboramarks 🧵

4

100

355

1

16

180

Yarin

@yaringal

4 years

Can autonomous 🚘 identify, recover from, and adapt to distribution shifts? We play with BDL and robust control to get cars to recover&adapt when they don't know what to do At ICML with @filangelos @ptigas @rowantmc @nick_rhinehart @svlevine 📄🎞️🕸️💻:

3

38

173

Yarin

@yaringal

3 years

A new class of attention-based models which learn _datasets_ instead of datapoints, and are able to solve tasks that traditional supervised neural nets cannot. Great work by @OATML_Oxford graduate students @janundnik * & @neilbband *, @clarelyle , @AidanNGomez

Jannik Kossen

@janundnik

3 years

🗞New Paper🗞 🤖🧪Self-Attention Between Datapoints: Going Beyond Individual Input-Output Pairs in Deep Learning 🧪🤖 Huge thanks to @neilbband * as well as @clarelyle , @AidanNGomez , @tom_rainforth , @yaringal , and @OATML_Oxford ! Introducing 🚀Non-Parametric Transformers🚀 1/

7

145

648

2

23

169

Yarin

@yaringal

6 years

Bayesian deep learning workshop @NipsConference 2017 accepted papers list is now online

Bayesian Deep Learning Workshop | NeurIPS 2021

Bayesian Deep Learning Workshop at NeurIPS 2021 — Tuesday, December 14, 2021, Virtual.

bayesiandeeplearning.org

2

49

149

Yarin

@yaringal

9 years

New machine learning art-works that *don't* use deep learning http://t.co/ZNYIc6LtP6 #ExtrapolatedArt http://t.co/gIScm2EMkS

5

52

148

Yarin

@yaringal

7 years

Another new work: Concrete dropout, with example Keras code to optimise over the dropout probabilities

2

47

144

Yarin

@yaringal

4 years

Nice metaphor from Summers-Stay: “GPT is like an improv actor who has never left home and only read about the world in books. Like such an actor, when it doesn’t know something, it will just fake it. You wouldn’t trust an improv actor playing a doctor to give you medical advice.”

Gary Marcus

@GaryMarcus

4 years

GPT-3 is a better bullshit artist than its predecessor, but it's still a bullshit artist. an investigation, @techreview , co-authored with Ernest Davis.

43

325

924

3

17

145

Yarin

@yaringal

3 years

Join us Thursday next week to hear @DavidDuvenaud talk about Infinitely Deep Bayesian Neural Networks with Stochastic Differential Equations! Also if you want to advertise your research during the BDL socials, send your poster here by 6/12:

Bayesian Deep Learning Workshop | NeurIPS 2021

Bayesian Deep Learning Workshop at NeurIPS 2021 — Tuesday, December 14, 2021, Virtual.

bayesiandeeplearning.org

0

14

140

Yarin

@yaringal

4 years

#WiML2019 with @ylecun

3

10

140

Yarin

@yaringal

4 years

@OATML_Oxford student Lewis Smith wrote a really interesting blogpost exploring his experience working with capsule networks -- explaining how to formulate a generative version of the model and how this revealed conceptual issues with capsules as a whole

4

29

140

Yarin

@yaringal

4 years

Hey @V7Labs can you send us a deck of cards?

7

29

138

Yarin

@yaringal

2 years

We have a fully funded PhD studentship to join @OATML_Oxford to work on systematic generalisation in ML, co-supervised between me and @egrefen . You'll get industry salary, spend 50% of your time at @UniofOxford and 50% at @facebookai (FAIR), with access to lots of compute etc ⬇️

Edward Grefenstette

@egrefen

3 years

⚠️ APPLICATION PROCESS ⚠️ Apply by emailing a CV, personal statement, and research proposal to oxford-fair-generalization-2022 @googlegroups .com by 📅 Nov 30 📅 (any time). Indicate if you would like Prof Foerster or Gal as your primary supervisor. 5/9

1

2

28

2

39

137

Yarin

@yaringal

3 years

I'm really excited to share 13 papers at @NeurIPSConf 2021 main conference and benchmarks track by @OATML_Oxford group members and collaborators!

13 OATML Conference papers at NeurIPS 2021 - OATML

OATML group members and collaborators are proud to present 13 papers at NeurIPS 2021 main conference.

oatml.cs.ox.ac.uk

0

13

134

Yarin

@yaringal

3 years

Many thanks to all my students and collaborators!

Oxford Comp Sci

@CompSciOxford

3 years

Many congratulations to Professor @yaringal who has been announced as one of 5 ‘Samsung AI Researcher of the Year’ award winners. Read more here:

0

5

42

6

3

123

Yarin

@yaringal

4 years

I often get questions about what it’s like to do an undergrad at Oxford @CompSciOxford . I didn’t do my own undergrad here, but I do find it a lot of fun teaching here (we have great students!) This thread is to give you some inside info on @UniofOxford if you plan to apply [1/n]

2

21

118

Yarin

@yaringal

7 years

What uncertainties do we need in Bayesian Deep Learning for Computer Vision? philosophy applied in the real world :)

2

50

118

Yarin

@yaringal

7 years

Many thanks to @nvidia and @AlisonBLowndes for the awesome GPU donation - expect to see lots more Bayesian neural net results out soon!

3

7

116

Yarin

@yaringal

5 years

#BDL Schedule, Accepted Papers, Contributed Talks, and Awards are updated online: Congrats to everyone who will be presenting at the workshop (136 abstracts accepted!). I'm looking forward to it

Bayesian Deep Learning Workshop | NeurIPS 2021

Bayesian Deep Learning Workshop at NeurIPS 2021 — Tuesday, December 14, 2021, Virtual.

bayesiandeeplearning.org

3

25

117

Yarin

@yaringal

5 years

A brilliant way to introduce ML to the general public (and to decision makers specifically) by Google Comics Factory. Making a good step towards answering a gap in education we discussed recently at @ESA_EO

2

23

113

Yarin

@yaringal

6 years

This made my day: "Bayesian deep learning" as a subject area for ICML submissions!

3

11

113

Yarin

@yaringal

4 years

I collected some stats from my @NeurIPSConf AC stack: * 66 reviewers * 46 engaged in discussion * 54 updated their review * 17 changed their score (highest delta was 3->6) * 119 discussion threads Curios to hear stats from other ACs / @NeurIPSConf itself

Dan Roy

@roydanroy

4 years

My faith in this part of the process is very limited. Requires both faithful reviewers and ACs. If either is missing, that's trouble.

3

0

9

10

3

111

Yarin

@yaringal

7 years

Code for Concrete Dropout now available online

0

24

108

Yarin

@yaringal

6 years

Have a look at , a digital magazine that aims to democratise research in AI. By @cnancyxu and others from Stanford

The Gradient

A digital publication about artificial intelligence and the future.

thegradient.pub

0

22

103

Yarin

@yaringal

4 years

Amazing work

What we learned from NeurIPS 2019 data

NeurIPS has quadrupled in the last five years. This year, we had 6,743 submissions after filtering (down to 6,614 at notification time)…

neuripsconf.medium.com

0

25

107

Yarin

@yaringal

4 years

Accelerating RL from rich observations, such as images, without relying on either domain knowledge or pixel-reconstruction: w/ Amy Zhang, @rowantmc , @RCalandra , @svlevine

Sergey Levine

@svlevine

4 years

Can we learn dynamics from images w/o reconstruction? By learning a latent space where distances obey bisimulation metric, we get latent states that group semantically similar but visually distinct states! w/ Amy Zhang, @rowantmc , @RCalandra , @yaringal

4

69

313

1

15

105

Yarin

@yaringal

8 months

Happy to share that I'll be helping the UK taskforce as director of research (together with @DavidSKrueger ). We're heavily recruiting - if you have technical expertise and want to work on Frontier Models (LLMs, generative AI), please read here

Ian Hogarth

@soundboy

8 months

@geoffreyhinton @ylecun @ericschmidt @sama 15/ @yaringal will join as Research Director of the Taskforce from Oxford where he is head of the Oxford Applied and Theoretical Machine Learning Group. Yarin is a globally recognised leader in Machine Learning.

2

33

2

17

106

Yarin

@yaringal

4 years

I want to see an example where GPT's output is "I don't know, ask an expert"

Yarin

@yaringal

4 years

Nice metaphor from Summers-Stay: “GPT is like an improv actor who has never left home and only read about the world in books. Like such an actor, when it doesn’t know something, it will just fake it. You wouldn’t trust an improv actor playing a doctor to give you medical advice.”

3

17

145

8

9

103

Yarin

@yaringal

5 years

Postbox is such a brilliant idea! Instead of continuously being bombarded by notifications, the app collects them and delivers them all together 3 times a day. Why isn't this an Android default?

Digital Wellbeing Experiments - Experiments with Google

Since 2009, coders have created thousands of amazing experiments using Chrome, Android, AI, WebVR, AR and more. We're showcasing projects here, along with helpful tools and resources, to inspire...

experiments.withgoogle.com

3

17

95

Yarin

@yaringal

4 years

We're glad to share 13 papers by @OATML_Oxford authors and collaborators to be presented at this @icmlconf and workshops. Full schedule here:

0

13

96

Yarin

@yaringal

4 years

This is the video we showed at the start of the Bayesian deep learning workshop panel discussion: @FinaleDoshi

ELBO

null

www.youtube.com

0

13

95

Yarin

@yaringal

4 years

This year @NeurIPSConf AC guidelines mention that the AC can "email the authors in the exceptional situation in which the [reviewers] discussion brings up new elements that would need to be clarified with the authors". This is very interesting - I wonder how many ACs will do this

2

4

93

Yarin

@yaringal

7 years

We're accepting submissions for the second Bayesian Deep Learning workshop at NIPS 2017:

0

41

92

Yarin

@yaringal

6 years

Long awaited work together with student @Adam_D_Cobb and Steve Roberts, on improving the safety of BDL models used in self-driving cars and in medical applications: "Loss-Calibrated Approximate Inference in Bayesian Neural Networks"

2

24

90

Yarin

@yaringal

1 month

twitter just told me that they've literally shadow banned me (reducing exposure of my posts) as punishment for not engaging enough with the platform I don't expect many people to see this...

11

2

91

Yarin

@yaringal

5 years

I really like the example distinguishing aleatoric from epistemic uncertainty "a lottery ticket (where the future, random outcome depends on chance) and a scratch card (where the outcome is already decided, but you don't know what it is)"

David Spiegelhalter

@d_spiegel

5 years

'Although numbers are often treated as cold, hard facts, we should be willing to acknowledge how uncertain they can be’. In my blog for Scientific American

7

76

179

3

13

89

Yarin

@yaringal

7 years

Code for our ICML paper "Dropout inference in BNNs with alpha-divergences" to identify adversarial images is online:

0

23

86

Yarin

@yaringal

5 years

A simple and robust tool for effective network pruning (few lines of code!). Just to make clear that this is work done by the great student @AidanNGomez ( @Deep__AI please at least cite the first author of the paper...)

DeepAI

@DeepAI

5 years

Great article published this week by @geoff_hinton @kswersk @yaringal about #NeuralNetwork #ComputerScience

0

9

22

2

9

88

Yarin

@yaringal

4 years

We are glad to share 23 papers by OATML authors and collaborators to be presented at this NeurIPS conference and workshops

3

4

89

Yarin

@yaringal

5 years

Another opportunity for a fully funded PhD in Oxford - ping me if you want to chat about applying to work with us @OATML_Oxford

Autonomous Intelligent Machines & Systems @Oxford

@aims_oxford

5 years

Fully-funded studentship with @aims_oxford @UniofOxford @oxbotica for entry in October 2020. Further details can be found at:

0

9

41

8

17

87

Yarin

@yaringal

1 year

Got a couple of new books for the lab @sirbayes Masashi Sugiyama

1

4

85

Yarin

@yaringal

6 years

All missing papers for the #BDL workshop are now available online

Bayesian Deep Learning Workshop | NeurIPS 2021

Bayesian Deep Learning Workshop at NeurIPS 2021 — Tuesday, December 14, 2021, Virtual.

bayesiandeeplearning.org

1

25

84

Yarin

@yaringal

6 years

Posters everywhere... I should probably start working on the talk :)

0

2

81

Yarin

@yaringal

3 years

It's been some time since I've written a blog post.. this one got a bit delayed due to NeurIPS SAC duties:

Why I Wouldn't Trust OpenAI's CLIP to Drive My Car - OATML

Massive models pre-trained on web-scraped data might be harmful to your downstream application, since their robustness is tightly tied to the way their data was curated.

oatml.cs.ox.ac.uk

1

12

83

Yarin

@yaringal

3 years

Party in D8

5

2

82

Yarin

@yaringal

3 years

Deep Kernel Learning combines GPs and deep learning in a principled way, keeping GPs' awesome properties But sometimes its uncertainty fails.. We investigated why - turns out it's Feature Collapse! Using recent tools we fix it to get principled Single Forward Pass Uncertainty

Joost van Amersfoort

@joost_v_amersf

3 years

Excited to share our work on single forward pass uncertainty for classification *and* regression! "On Feature Collapse and Deep Kernel Learning for Single Forward Pass Uncertainty" 👉 Simple & extensible implementation: Summary:👇

1

32

174

0

8

80

Yarin

@yaringal

6 years

"There comes a time where you'll know so many things that as you forget them, you can reconstruct them from the pieces that you can still remember. It is therefore of first-rate importance that you know how to triangulate - figure something out from what you already know" Feynman

0

14

76

Yarin

@yaringal

7 years

New work on real-time image saliency from my master's student Piotr Dabkowski - great job!

0

23

78

Yarin

@yaringal

6 years

Countdowns to top CV/NLP/ML/Robotics/AI conference deadlines #machinelearning

0

31

78

Yarin

@yaringal

7 years

Bayesian Deep Learning @NipsConference workshop schedule is online

Bayesian Deep Learning Workshop | NeurIPS 2021

Bayesian Deep Learning Workshop at NeurIPS 2021 — Tuesday, December 14, 2021, Virtual.

bayesiandeeplearning.org

3

30

74

Yarin

@yaringal

3 years

PhD opportunity with us at @OATML_Oxford funded by @facebookai , co-supervised between me and @egrefen to work on systematic generalisation. Take a look at the thread if interested / share to others who might be interested

Edward Grefenstette

@egrefen

3 years

🧵THREAD 🧵 Are you looking to do a 4 year Industry/Academia PhD? I am looking for 1 student to pioneer our new FAIR-Oxford PhD programme, spending 50% if their time at @UniofOxford , and 50% at @facebookai (FAIR) while completing a DPhil (Oxford PhD). Interested? Read on… 1/9

10

113

483

2

14

75

Yarin

@yaringal

3 years

Did you like @dustinvtran 's @NeurIPSConf tutorial on Uncertainty in Deep Learning? join us *tomorrow at 11am GMT* for the NeurIPS Europe meetup on Bayesian Deep Learning, co-organised with ELLIS, to find out about the latest research in the field!

Bayesian Deep Learning Workshop | NeurIPS 2021

Bayesian Deep Learning Workshop at NeurIPS 2021 — Tuesday, December 14, 2021, Virtual.

bayesiandeeplearning.org

Dustin Tran

@dustinvtran

3 years

Snippet 26: Open challenge of benchmarks. Announcing Uncertainty Baselines lead by @zacharynado to easily build upon well-tuned methods! Joint effort with @OATML_Oxford

1

10

26

1

6

71

Yarin

@yaringal

5 years

Excited to be speaking at the UN's "AI for Good" Global Summit this Wednesday! I'll be talking about our work with @NASA @nasa_fdl and @esa . Also looking forward to meet the formidable collection of speakers at the summit: @OATML_Oxford @UniofOxford

3

6

71

Yarin

@yaringal

6 years

AI Safety Gridworlds, from DeepMind "We present a suite of RL envs illustrating safety properties of intelligent agents [..] We evaluate A2C and Rainbow on our envs and show that they are not able to solve them satisfactorily" We need more papers like this

AI Safety Gridworlds

We present a suite of reinforcement learning environments illustrating various safety properties of intelligent agents. These problems include safe interruptibility, avoiding side effects, absent...

arxiv.org

0

15

73

Yarin

@yaringal

7 years

Zoubin's Keynote talk at the Bayesian Deep Learning Workshop at #NIPS2016 is available as well now!

History of Bayesian Neural Networks (Keynote talk)

Zoubin Ghahramani (University of Cambridge)---Bayesian Deep Learning WorkshopNIPS 2016December 10, 2016 — Centre Convencions Internacional Barcelona, Barcelo...

www.youtube.com

0

26

72

Yarin

@yaringal

8 years

We're organising a Bayesian deep learning workshop at NIPS 2016 CFP online

Bayesian Deep Learning Workshop | NeurIPS 2021

Bayesian Deep Learning Workshop at NeurIPS 2021 — Tuesday, December 14, 2021, Virtual.

bayesiandeeplearning.org

1

39

72

Yarin

@yaringal

4 years

Work by many collaborators (including three @OATML_Oxford members) modelling the effect of different non-pharmaceutical interventions against #COVID19 transmission, with extensive empirical validation. Great job Jan Brauner, @sorenmind , and everyone!

3

20

71

Yarin

@yaringal

3 years

We're glad to share 22 papers by @OATML_Oxford authors and collaborators to be presented at this @NeurIPSConf and workshops. Full schedule here:

0

9

70

Yarin

@yaringal

4 years

I disagree. It is our responsibility as ACs to ensure reviewers read the rebuttal and defend their decision. A reviewer ignores author rebuttal? AC should report them so they are not invited to review again. But many ACs don't engage.. We need mechanisms to highlight that to SACs

Dan Roy

@roydanroy

4 years

Absolutely!

0

9

8

4

69

Yarin

@yaringal

4 years

Christ Church college, Oxford, is advertising a Junior Research Fellowship (JRF) in Computer Science, tenable from 1 Oct 2021: JRFs are offered by Oxford colleges to early career researchers to develop their independent research Deadline 20h November 2020

1

14

67

Yarin

@yaringal

4 years

I remember thinking "what can I do to help". Then I realised that I'm lucky to actually be in a very powerful position--doing admissions at @UniofOxford . I used to complain that there's not enough diversity. Now I work to _create_ diversity. Guys, instead of complaining, act.

Katherine Heller

@kat_heller

4 years

@fhuszar Your wealthy tech/liberal bubble. In which how many people are black? (Or even women?) In which how many of us do more than tweet about these issues? We want to believe that people can look to us for social progress. But they can’t. Change starts with ourselves.

1

2

11

5

1

68

Yarin

@yaringal

8 years

I put the code from Dropout Uncertainty paper on Github. Playing w it & setting 10x epochs got big RMSE improvement!

GitHub - yaringal/DropoutUncertaintyExps: Experiments used in "Dropout as a Bayesian Approximation:...

Experiments used in "Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning" - yaringal/DropoutUncertaintyExps

github.com

0

19

64

Yarin

@yaringal

7 years

Bayesian Deep Learning workshop @NipsConference CFP: Deadline is *this Friday 3 November* ICLR submissions welcomed!

Bayesian Deep Learning Workshop | NeurIPS 2021

Bayesian Deep Learning Workshop at NeurIPS 2021 — Tuesday, December 14, 2021, Virtual.

bayesiandeeplearning.org

3

11

66

Yarin

@yaringal

4 years

We have a fully funded PhD studentship with the Satellite Application Catapult and Deimos Space UK, to work on AI for Good & Earth Observations (EO), or to develop new ML methodology towards EO. Deimos has some amazing challenges in AI for Good & EO [1/2]

2

15

64

Yarin

@yaringal

6 years

'“Is curing patients a sustainable business model?” Goldman Sachs analysts ask' - sigh

“Is curing patients a sustainable business model?�� Goldman Sachs analysts ask

Analyst report notes that Gilead’s hep C cure will make less than $4 billion this year.

arstechnica.com

7

22

64

Yarin

@yaringal

4 years

This is not a normal #neurips year. We've had disruptions on a global scale and many have been disadvantaged by these. I'll email the PC advocating for @roydanroy 's suggestion to offer extensions to anyone who's been personally affected and willing to write to them CC @kat_heller

5

9

63

Yarin

@yaringal

3 years

Have you been wondering what @dpkingma has been up-to over the past couple of years? The BDL schedule is now online: Registration for the event is free, but we can only fit limited numbers so make sure to register early:

Bayesian Deep Learning Workshop | NeurIPS 2021

Bayesian Deep Learning Workshop at NeurIPS 2021 — Tuesday, December 14, 2021, Virtual.

bayesiandeeplearning.org

2

8

61

Yarin

@yaringal

5 years

Have a look at our latest @OATML_Oxford blog post by @ZacKenton1 , @filangelos and Owain Evans, studying how generalisation interacts with safety in RL:

1

17

60

Yarin

@yaringal

6 years

It's been online for a few days, but still: Continual learning is important for medical applications, yet evals in the field completely ignore the point of learning continually! worse, with more sensible evals existing algos fail.. gap in research for u :)

Towards Robust Evaluations of Continual Learning

Experiments used in current continual learning research do not faithfully assess fundamental challenges of learning continually. Instead of assessing performance on challenging and representative...

arxiv.org

2

16

60

Yarin

@yaringal

7 years

Congrats to the authors of the 41 accepted abstracts to the NIPS 2016 Workshop on Bayesian Deep Learning!

Bayesian Deep Learning Workshop | NeurIPS 2021

Bayesian Deep Learning Workshop at NeurIPS 2021 — Tuesday, December 14, 2021, Virtual.

bayesiandeeplearning.org

3

21

59

Yarin

@yaringal

5 years

Apply now to this fully funded PhD in ML! You can work with lots of great PIs across Oxford and Imperial, including @OATML_Oxford :)

StatML CDT

@StatMLIO

5 years

PhD applications are now open for the EPSRC CDT in Modern Statistics and Statistical Machine Learning at Imperial and Oxford for October 2020 starts. Apply now!

0

37

89

0

11

59

Yarin

@yaringal

6 years

"gcc -O4 sends your code to Jeff Dean for a complete rewrite" :) Are there any "Jeff Dean ML facts"?

3

11

59

Yarin

@yaringal

4 years

Very well put @andrewgwils . "BDL is gaining visibility because we are making progress. We shouldn’t discourage these efforts. If we are shying away from an approximate Bayesian approach because of some challenge or imperfection, we should always ask, “what’s the alternative”?"

Andrew Gordon Wilson

@andrewgwils

4 years

Bayesian methods are *especially* compelling for deep neural networks. The key distinguishing property of a Bayesian approach is marginalization instead of optimization, not the prior, or Bayes rule. This difference will be greatest for underspecified models like DNNs. 1/18

16

384

2K

0

4

59

Yarin

@yaringal

5 years

Ping me if you want to join us for a PhD at @OATML_Oxford under the AIMS programme (fully funded PhD). Unrelated, also opportunities to work on ML projects with @esa

Michael A Osborne

@maosbot

5 years

AIMS is accepting applications for fully-funded @UniofOxford PhD places to work on machine learning, vision, robotics, sensor networks and more. The deadline is **25 Jan**. More details and the application site: Please spread the word!

2

49

94

6

7

59

Yarin

@yaringal

5 years

Thank you all for coming to the #BDL workshop @NipsConference 2018, and thank you so much to my wonderful co-organisers Christos Louizos, Miguel Hernández-Lobato, Andrew G. Wilson, Zoubin Ghahramani, Kevin Murphy, and Max Welling @wellingmax @andrewgwils @ChrLouizos

2

7

58

Yarin

@yaringal

3 years

Congratulations to @OATML_Oxford graduate students @filangelos and @timrudner for receiving a 2021 J.P. Morgan PhD Fellowship and a 2021 Qualcomm Innovation Fellowship!

3

4

58

Yarin

@yaringal

5 years

We've added a summary of reviewers' points and rebuttals for our ICLR rejected paper, for the convenience of new readers: The paper sets the scene for future research into robustness to adv examples in BNNs, and gives lots of insights into what's going on

Sufficient Conditions for Robustness to Adversarial Examples: a...

We prove that idealised Bayesian neural networks can have no adversarial examples, and give empirical evidence with real-world BNNs.

openreview.net

2

9

56

Yarin

@yaringal

6 years

Understanding measures of uncertainty for adversarial example detection - great job by Lewis Smith!

1

16

55