Umang Bhatt @umangsbhatt Twitter profile | Pikagi

Pikagi

Umang Bhatt

@umangsbhatt

1,567

Followers

327

Following

34

Media

477

Statuses

ML Faculty @NYUDataScience . Research @TuringInst . Previously: PhD @Cambridge_Uni , @Mozilla , @PartnershipAI , @CarnegieMellon .

🇺🇸⇔🇬🇧

https://t.co/FighiAPtUp

Joined July 2016

Don't wanna be here? Send us removal request.

Pinned Tweet

@umangsbhatt

Umang Bhatt

1 year

I am joining @NYUDataScience as an Assistant Professor/Faculty Fellow this fall! I look forward to continuing research in trustworthy #machinelearning and building #AI systems that augment and complement human decision-makers 🚀

@NYUDataScience

NYU Data Science

@NYUDataScience

1 year

Meet CDS Assistant Professor/Faculty Fellow Umang Bhatt ( @umangsbhatt ), who will join CDS this fall. Umang will pursue research in trustworthy #machinelearning , #ResponsibleAI , and human-machine collaboration at NYU. More on today's blog!

Tweet media one

1

3

48

32

8

252

Last Seen Profiles

@yoko_harada_619

@system_chaser

@dwrkblink

@Roma_una_

@iamr_e_al

@031002_sh

@hiimmikelee

@HelouRasha

@turbanlisever69

@ZRawwest

@AKB_1217

@AniitaDraw

@congsime

@HK416D__

@HardstuckYT

@taNTRum_guy

@Peronismox100

@G2NiKo

@updatesCLC

@deeryeo

@Cryptodiamond20

@CYN7HI4_

@marioalex556594

@bostonfitnew

@MGorashova

@Toastfnfn

@glisten_ai

@RYOTA_Equal

@erroryuu

@Macosta2547

@BrawlStars

@Goddess__Tawny

@26wrldtae

@NasserPolly

@Crypto4gemspace

@N1info

@umangsbhatt

Umang Bhatt

7 months

I successfully defended my PhD thesis today 🎉 Many thanks to my examiners, @erichorvitz and @lawrennd , for their thoughtful comments and a very fun viva!

Tweet media one

Tweet media two

62

24

633

@umangsbhatt

Umang Bhatt

3 years

Happy to share our review paper: “Uncertainty as a Form of Transparency: Measuring, Communicating, and Using Uncertainty” Draft: 🧵👇

Tweet card media

Uncertainty as a Form of Transparency: Measuring, Communicating,...

Algorithmic transparency entails exposing system properties to various stakeholders for purposes that include understanding, improving, and contesting predictions. Until now, most research into...

3

28

164

@umangsbhatt

Umang Bhatt

2 years

Announcing the 2022 @icmlconf Workshop on Human-Machine Collaboration and Teaming! We are interested in the algorithmic and socio-computational paradigms needed to design and support human-machine collaboration. Do consider submitting! Deadline: May 16

4

38

147

@umangsbhatt

Umang Bhatt

4 years

Excited to be a @mozilla fellow for the next year! 🦊 Thankful for their support in our work towards trustworthy and transparent machine learning 😀

6

8

144

@umangsbhatt

Umang Bhatt

3 years

Excited for this new adventure! Looking forward to new collaborations and more frequent London trips @turinginst

@Cambridge_Eng

Engineering Dept

3 years

PhD student & Mozilla Fellow Umang Bhatt has been accepted into the Turing Enrichment Scheme at the Alan Turing Institute. Superviser @adrian_weller @umangsbhatt @turinginst @CambridgeMLG #TuringEnrichment #machinelearning #ArtificialIntelligence

Tweet media one

0

6

24

4

1

88

@umangsbhatt

Umang Bhatt

3 years

Excited for our @ELLISforEurope Workshop on Human-Centric Machine Learning on Monday, May 10th from 3pm to 9pm UK Time (7am to 1pm Pacific Time) Register to attend at

1

27

67

@umangsbhatt

Umang Bhatt

2 years

Spending the next few months in the *other* Cambridge at @HCRCS and @HSEAS , visiting @hima_lakkaraju and @MilindTambe_AI Happy to chat if you’re in the Boston area! I’m looking for nearby coffee recs too ☕️

Tweet media one

Tweet media two

Tweet media three

7

0

63

@umangsbhatt

Umang Bhatt

2 years

If you're at #ICML2022 , check out our poster “Perspectives on Incorporating Expert Feedback into Model Updates” at the Updatable ML workshop in Ballroom 2 tomorrow at 3:30pm. 🧵 Joint w/ @valeriechen_ , @HodaHeidari , @adrian_weller , @atalwalkar Preprint:

Tweet media one

4

12

59

@umangsbhatt

Umang Bhatt

5 years

Excited that our paper on how various explainability algorithms are used in practice () was accepted to #FAT2020 @fatconference 🎉

Tweet card media

Explainable Machine Learning in Deployment

Explainable machine learning offers the potential to provide stakeholders with insights into model behavior by using various methods such as feature importance scores, counterfactual explanations,...

1

8

59

@umangsbhatt

Umang Bhatt

10 months

Heading to Accra for my first @DeepIndaba ! Excited to explore Ghana 🇬🇭 Do reach out if you’ll be around

4

6

51

@umangsbhatt

Umang Bhatt

2 years

🎉 Thanks to @adrian_weller , @HMAArnold , and @stephenjcave for their unending support through my PhD thus far!

@Cambridge_Eng

Engineering Dept

2 years

PhD student @umangsbhatt awarded a @jpmorgan AI #PhD #Fellowship Umang works to ensure that machine learning technologies are developed responsibly. As #AI makes ever more decisions in our lives, his research is incredibly important. @StEdmundsCam

Tweet media one

1

6

46

3

3

52

@umangsbhatt

Umang Bhatt

4 years

My collaborator, Botty Dimanov, presenting recent work on concealing model unfairness from explanation methods #AAAI20 TLDR: Don’t use feature importance to justify the discriminatory nature of a model.

Tweet media one

1

3

29

@umangsbhatt

Umang Bhatt

4 years

Excited to have @zacharylipton visiting @CambridgeMLG today

Tweet media one

1

1

29

@umangsbhatt

Umang Bhatt

4 years

Only *two* weeks left to submit to the @icmlconf Workshop on Human Interpretability. Submit your latest work on interpretability, explainability, and transparency #ICML2020

@umangsbhatt

Umang Bhatt

4 years

In non-COVID news, the CFP for the 5th annual @icmlconf Workshop on Human Interpretability in Machine Learning ( #WHI2020 ) is now live 🙃 Consider submitting your work on interpretability and XAI; this year’s focus will be “Intepretability in Practice”

0

2

26

1

11

29

@umangsbhatt

Umang Bhatt

4 years

Excited to share a preprint of our forthcoming #IJCAI20 paper: "Evaluating and Aggregating Feature-based Model Explanations" Joint work with @adrian_weller and @josemfmoura (1/3)

3

3

28

@umangsbhatt

Umang Bhatt

5 years

. @hannawallach vehemently urging us not to forget ML is full of value-laden decisions, just like other disciplines — remember this while trying to build human-centric, “intelligible” systems 🔥 #NeurIPS2019

0

4

27

@umangsbhatt

Umang Bhatt

4 years

In non-COVID news, the CFP for the 5th annual @icmlconf Workshop on Human Interpretability in Machine Learning ( #WHI2020 ) is now live 🙃 Consider submitting your work on interpretability and XAI; this year’s focus will be “Intepretability in Practice”

Overview The Fifth Annual Workshop on Human Interpretability in Machine Learning (WHI 2020), held in conjunction with ICML 2020, will bring together artificial intelligence (AI) researchers who study...

sites.google.com

0

2

26

@umangsbhatt

Umang Bhatt

6 years

After a hint of this and that turning me black, Swami has given me my true colors back #mokshanow #nc18baps

0

3

23

@umangsbhatt

Umang Bhatt

3 years

Attending @AIESConf ? Come talk to us about this work and about uncertainty more generally. We’ll be at poster sessions 2 and 3 #AIES2021

@umangsbhatt

Umang Bhatt

3 years

Happy to share our review paper: “Uncertainty as a Form of Transparency: Measuring, Communicating, and Using Uncertainty” Draft: 🧵👇

3

28

164

1

5

25

@umangsbhatt

Umang Bhatt

3 years

Quite excited about our work on uncertainty explanations, which we call CLUEs 🔍 To appear at #ICLR2021 as an oral presentation @iclr_conf @trustworthy_ml

Tweet card media

Getting a CLUE: A Method for Explaining Uncertainty Estimates

Both uncertainty estimation and interpretability are important factors for trustworthy machine learning systems. However, there is little work at the intersection of these two areas. We address...

@JaviAC7

Javier Antorán

3 years

Our CLUE paper was accepted to #ICLR2021 as an oral presentation! Big thanks to my collaborators @umangsbhatt , @tameem_adel , @adrian_weller , my supervisor @jmhernandez233 and the reviewers, who provided great feedback. A draft of the paper can be found at

2

7

40

0

3

24

@umangsbhatt

Umang Bhatt

4 years

Sadly not in Barcelona for #ICASSP2020 . I will be presenting our paper "On Network Science and Mutual Information for Explaining Neural Networks" remotely this Friday! w/ Brian Davis, Kartikeya Bhardwaj, @rmarculescu @utexasece , @josemfmoura @CMU_ECE

Tweet card media

On Network Science and Mutual Information for Explaining Deep...

In this paper, we present a new approach to interpret deep learning models. By coupling mutual information with network science, we explore how information flows through feedforward networks. We...

1

1

23

@umangsbhatt

Umang Bhatt

4 years

We've reached THAT point near a deadline where my collaborator messages me: I added the sentence "But, what are probabilities?"

3

0

22

@umangsbhatt

Umang Bhatt

2 years

Very excited about our excellent speakers/panelists: Sir Tim Gowers ( @wtgowers ), Wendy Mackay, Ernest Mwebaze ( @emwebaze ), Nuria Oliver ( @nuriaoliver ), Phebe Vayanos ( @phebe_vayanos ), Petar Veličković ( @PetarV_93 ), Fernanda Viégas ( @viegasf ), and Martin Wattenberg ( @wattenberg )

Tweet media one

1

3

21

@umangsbhatt

Umang Bhatt

7 months

So many people to thank, but my guru, parents, sister @naitisb , and advisor @adrian_weller have been exceptionally supportive over the years 🙏🏾 @Cambridge_Eng , @CambridgeMLG @LeverhulmeCFI , and @StEdmundsCam provided a wonderful environment for my PhD!

3

0

20

@umangsbhatt

Umang Bhatt

8 months

Nice to attend #EAAMO23 where we proposed "FeedbackLogs" led by Matthew Barker and @EmmaKallina We provide a new perspective on documentation that holds practitioners accountable for actioning upon stakeholder feedback when building AI systems @ACMEAAMO

Tweet media one

1

3

20

@umangsbhatt

Umang Bhatt

3 years

Sharing an RA opportunity to collaborate on explainability w/ Frens Kroger, Beate Grawemeyer, @jeffhancock , and me. We're looking to study trust formation via transparency and its relation to online harms.

1

9

16

@umangsbhatt

Umang Bhatt

2 years

🙌🏽

@BirkhoffGuo

Hangzhi Guo

2 years

Delighted to host @umangsbhatt (Ph.D. candidate @CambridgeMLG ) as our next @PennStateCSRAI Young Achievers Symposium. He will deliver a talk about explainable machine learning #XAI on April 12th, 4 - 5 pm.

Tweet media one

1

0

7

0

2

17

@umangsbhatt

Umang Bhatt

1 year

Excited about our working paper on human-centered, *interactive* evaluation of LLMs as assistants for mathematicians. Room for future work on how beginners and experts alike may start to use LLMs appropriately and selectively in their existing workflows!

@katie_m_collins

Katie Collins

@katie_m_collins

1 year

Evaluating large language models is hard, particularly for mathematics. To better understand LLMs, it makes sense to harness *interactivity* - beyond static benchmarks. Excited to share a new working paper ! 1/

Tweet media one

5

88

451

1

1

17

@umangsbhatt

Umang Bhatt

4 years

My precious (late) laptop reappeared online in Morocco last night. May the force be with the new owner ✊🏽

Tweet media one

@umangsbhatt

Umang Bhatt

4 years

Love-hate relationship with Barcelona. PROs: experiencing a mix of urban grunge and rich Spanish history was refreshing; plus @FAccTConference was lovely CONs: my laptop bag (and its contents) found a new home, thanks to ladrones. I hope the new owners enjoy my (ex-)machine 😭

1

0

7

0

0

16

@umangsbhatt

Umang Bhatt

2 years

We have extended the submission deadline for our ICML Workshop on Human-Machine Collaboration and Teaming to May 26! Consider submitting any recent or in-progress human-AI partnership work, including your @NeurIPSConf submissions or your @acm_chi / @FAccTConference papers :)

@umangsbhatt

Umang Bhatt

2 years

Announcing the 2022 @icmlconf Workshop on Human-Machine Collaboration and Teaming! We are interested in the algorithmic and socio-computational paradigms needed to design and support human-machine collaboration. Do consider submitting! Deadline: May 16

4

38

147

0

6

16

@umangsbhatt

Umang Bhatt

5 years

Thrilled to be in SF this summer working on exciting, multi-stakeholder problems in AI

@PartnershipAI

Partnership on AI

5 years

@umangsbhatt spearheads our work in AI #explainability - including quantitative evaluation criteria for feature-level explanations and understanding how information flows through deep neural networks.

Tweet media one

1

1

11

0

1

16

@umangsbhatt

Umang Bhatt

4 years

Come join us! We want to study how to deploy explainability for transparency goals #XAI

@alicexiang

Alice Xiang

4 years

@PartnershipAI is hiring for a Research Fellow in Explainable ML! Our goal is to study how XAI can better enable transparency, taking into account the nuances of context and the needs of different stakeholders.

1

5

14

0

2

15

@umangsbhatt

Umang Bhatt

4 years

Saliency maps help build trust

@medburnbook

status annoyicus

4 years

give me a horror story from your specialty in five words or less

19K

867

9K

0

2

15

@umangsbhatt

Umang Bhatt

11 months

Glad our review is now published. It organizes our thinking on how to leverage human feedback and domain expertise at different stages of learning ML models Joint w/ @valeriechen_ @HodaHeidari @adrian_weller @atalwalkar

Tweet card media

Perspectives on incorporating expert feedback into model updates

We review techniques that enable practitioners to incorporate feedback from domain experts into updates to machine learning models. Our work motivates the development of new methods for soliciting...

@Patterns_CP

Patterns journal

11 months

Perspectives on incorporating expert feedback into model updates @valeriechen_ and co-authors review in detail methods for updating #machinelearning & #AI models with expert knowledge, and propose a taxonomy for categorizing different feedback types

0

1

6

1

1

14

@umangsbhatt

Umang Bhatt

4 years

Mozilla has been a champion in making systems open and accessible. As they begin establishing their trustworthy AI agenda, algorithmic transparency (explainability, uncertainty, etc.) is a key piece of the puzzle. Looking forward to the new collaborations this brings!

2

0

14

@umangsbhatt

Umang Bhatt

5 years

Entertaining to hear @zacharylipton talk to a room of artists and creative ML folks at @mldcmu -- "interpretability is the catch-all word for all things humans want from ML models": it's basically an antelope, your local waste-basket taxon (aka the growing dumpster fire out back)

1

2

14

@umangsbhatt

Umang Bhatt

4 years

The #ICML2020 Workshop on Human Interpretability in ML has begun! Check out the papers here: Huge thanks to @hen_str for helping put this virtual experience together! @trustworthy_ml

1

4

13

@umangsbhatt

Umang Bhatt

1 year

Cambridge has been a dream 🙌🏾Can’t wait for my final months as a student @CambridgeMLG @LeverhulmeCFI , but I’m stoked to move back to 🇺🇸 and live in NYC 🗽 Many thanks to my advisor @adrian_weller ! I look forward to more collaboration, adventure, and learning for years to come

0

0

13

@umangsbhatt

Umang Bhatt

4 years

Looking forward to speaking at this with @JaviAC7

@PartnershipAI

Partnership on AI

4 years

While there are countless articles on methods and techniques that could be used to enable #explainability , how are these techniques actually used in deployed systems? Join PAI Research Fellow @umangsbhatt this Wednesday as he presents his latest research:

0

9

21

1

1

13

@umangsbhatt

Umang Bhatt

1 year

The potential for diverse collaborations within @NYUDataScience (w/ @stoyanoj , @LakeBrenden , @sleepinyourhat , @andrewgwils ) and the @nyuniversity ecosystem (w/ the likes of @merbroussard , @JonHaidt , @Lawgeek , @weijima01 ) make CDS an exciting place for responsible AI research 🤝🏾

2

0

13

@umangsbhatt

Umang Bhatt

4 years

Come join @CambridgeMLG -- tons of exciting, interdisciplinary directions to explore wrt trustworthy ML! @trustworthy_ml

@adrian_weller

Adrian Weller

4 years

PhD positions in Machine Learning at Cambridge. Please apply by Dec 3! Opportunities to work with me on trustworthy ML including fairness, interpretability and robustness. Feel free to get in touch if you have questions.

2

61

201

0

1

13

@umangsbhatt

Umang Bhatt

5 years

Life update: Excited to have joined @PartnershipAI as a Postgraduate Research Fellow today! Looking forward to a summer full of explainability and responsible AI.

1

0

13

@umangsbhatt

Umang Bhatt

4 years

Cool to see @azeem , @libbykinsey , and co highlight our explainability in practice work @PartnershipAI

@ExponentialView

Exponential View 🔮

@ExponentialView

4 years

Breakthroughs in #AI come at an ever more rapid pace. We are building a bridge between the newest research and its relevance for businesses, product leaders and data science teams across the world in this new report➡️

Tweet media one

0

2

8

0

3

12

@umangsbhatt

Umang Bhatt

2 years

Might be of interest @trustworthy_ml @HumanCenteredAI @black_in_ai @_LXAI @AiDisability @wimlds

0

1

12

@umangsbhatt

Umang Bhatt

2 years

Apply to join us @CambridgeMLG ! It’s a wonderful place to pursue research w/ @adrian_weller and others

@adrian_weller

Adrian Weller

2 years

Apply for a phd in machine learning at @CambridgeMLG to work with me or wonderful colleagues. I have broad interests, mostly centred on @trustworthy_ml that can be deployed at scale. *US citizens apply by Oct 12 to be eligible for a Gates scholarship*

1

20

66

0

1

12

@umangsbhatt

Umang Bhatt

4 years

Loving this summary by @ecekamar on work with @bansalg_ , @dsweld , and @erichorvitz regarding backwards compatibility, dissonance in model errors, and designing for complementarity #SafeAI2020 #AAAI20 Also: good morning, New York :)

1

1

11

@umangsbhatt

Umang Bhatt

6 years

What a surreal experience meeting so many interdisciplinary researchers at #FAT2018 - everyone here is radiating a deep desire to understand algorithmic fairness, tackle model interpretability, and make a damn difference in this world

1

3

11

@umangsbhatt

Umang Bhatt

4 years

Nice to see LinkedIn using fair rank aggregation in search and deploying feature importance explanations for engineers during model selection; h/t @kkenthapadi in the @fiddlerlabs tutorial #FAT2020

0

0

11

@umangsbhatt

Umang Bhatt

5 years

Also thankful for those that shared thoughts and feedback with us: @hima_lakkaraju @krishnagade @bansalg_ @terahlyons @krvarshney @rajiinio @frossi_t @PeterLoPR @saayelimukherji @a_b_powell @KarinaAlexanyan @gabizij @pjturcot @ejette @RosieCampbell @nicole_rigillo

1

0

11

@umangsbhatt

Umang Bhatt

3 years

Incredible, heart-wrenching, and stimulating talk by @ruha9 this evening

Tweet media one

1

0

10

@umangsbhatt

Umang Bhatt

5 years

Interesting thoughts from @harari_yuval I read on the way to #AAAI19 and #AIES : “The real problem with robots [and models] is not their own artificial intelligence but the natural stupidity and cruelty of their human masters...”

0

0

10

@umangsbhatt

Umang Bhatt

4 years

Looking forward to a fun, productive week in Barcelona for the @fatconference

1

1

10

@umangsbhatt

Umang Bhatt

6 years

Wonderful day at the @Affectiva Emotion AI Summit on Trust in AI. Thanks for putting together an amazing lineup @kaliouby and thanks for the intro @navrina_singh - looking forward to building a human-machine social contract together!

2

2

10

@umangsbhatt

Umang Bhatt

3 years

This work was done jointly with @javiac7 Yunfeng Zhang @qveraliao @prasatti Riccardo Fogliato @gabriellegma @jstanl Ranganath Krishnan @omeshtickoo @lamanachman @rumichunara @madhusrikumar @alicexiang @adrian_weller

1

1

10

@umangsbhatt

Umang Bhatt

2 years

Co-organized with Katie Collins ( @katie_m_collins ), Maria De-Arteaga ( @mariadearteaga ), Bradley Love ( @profdata ), and Adrian Weller ( @adrian_weller )

1

1

10

@umangsbhatt

Umang Bhatt

6 years

Back in PGH after an awesome trip to LA and a wonderful event. Till next time @fennell_p @talamobile , @ToddTerrazas , @phebe_vayanos @CAIS_USC , @HealthcareWen @CATALAIZE , @ccansu @aiethicslab , Brent Venable @TulaneSSE @TheIHMC , @jkeltner @Upstart , @CGMikeHowells #AIEthics

Tweet media one

0

3

10

@umangsbhatt

Umang Bhatt

3 years

The hardest part was unifying the seemingly disparate literature and piecing together the varied language in each discipline. We wanted to understand how to effectively design and deploy uncertainty-aware ML models to maximally benefit practitioners and end-users.

2

1

9

@umangsbhatt

Umang Bhatt

10 months

I’ll be running a practical on Responsible AI with @KanubalaD @AishaAlaagib @myzta_tee on Monday (Sept 4) at 2pm. Looking forward to it! @KaliTessera has curated an awesome set of pracs

@DeepIndaba

Deep Learning Indaba

10 months

Have you had a chance to checkout all the exciting Practicals that will be happening at the Indaba? A huge shoutout goes out to all the teams that have been developing exciting code and content for all of us to learn from. Which one are you looking forward to the most?…

Tweet media one

Tweet media two

Tweet media three

0

24

60

0

3

9

@umangsbhatt

Umang Bhatt

5 years

Maybe every preprint should come with a video "to catapult immature work into the public view" -- can someone at @wef and/or @KayFButterfield please read @zacharylipton 's sobering piece on OpenAI instead of stringing together apocalyptic phrases to cue Doomsday

2

1

9

@umangsbhatt

Umang Bhatt

3 years

Our workshop touches upon two themes: 1. The differential treatment by algorithms of historically under-served and disadvantaged communities 2. The development of machine learning systems to assist humans for better performance, rather than replace them

1

1

8

@umangsbhatt

Umang Bhatt

6 years

Extremely proud of my close friends @oliviakoshy and @malharhar for recently presenting their AI & Children’s Rights research to the folks at @wef - can’t wait to see what problems they tackle next ❤️ #humanizeAI @hrcberkeley @hrdag

Tweet media one

0

2

8

@umangsbhatt

Umang Bhatt

3 years

A few months later, @viegasf pointed us to the work of @jessicahullman -- what a gold mine! So now we were in an abyss between the Bayesian approaches to ML and absolutely beautiful depictions of uncertainty (seriously just check out anything from @adamrpearce ).

1

1

8

@umangsbhatt

Umang Bhatt

3 years

Looking forward to this later today with @enablingNGOlaw and @COE4AI #MozFest #CAHAI #Transparency

@enablingNGOlaw

ECNL

@enablingNGOlaw

3 years

🎉Another session we are hosting at #MozFest ? Join our @vanjasko leading a discussion w/ @umangsbhatt & representative from @COE4AI #CAHAI on #participation in the #AI regulation debate. ⏰When: March 19, 14:30 CET 🖱️ Sign up here:

Tweet media one

1

3

9

0

2

8

@umangsbhatt

Umang Bhatt

5 years

Even more thrilled since this is my first paper since joining @CambridgeMLG @LeverhulmeCFI as a doctoral student and joining @PartnershipAI as a research fellow #phdchat @AcademicChatter 😃

0

0

7

@umangsbhatt

Umang Bhatt

5 years

When visiting @AmiBhattMD and @neilmaniar recently, I learned about Interface Age (a "computer" magazine running from 1976 to 1984). What a hidden gem!

Tweet media one

1

0

7

@umangsbhatt

Umang Bhatt

8 months

Paper: Summary: Joint work with Matthew Barker, @EmmaKallina , Dhananjay Ashok, @katie_m_collins , @AshleyCasovan , @adrian_weller , @atalwalkar , and @valeriechen_ @turinginst @ResponsibleAI @LeverhulmeCFI @CambridgeMLG @mldcmu

Tweet card media

FeedbackLogs: Recording and Incorporating Stakeholder Feedback into Machine Learning Pipelines |...

🔬 Research Summary by Matthew Barker, a recent graduate from the University of Cambridge, whose research focuses on explainable AI and human-machine teams. [Original paper by Matthew Barker…

montrealethics.ai

0

0

7

@umangsbhatt

Umang Bhatt

4 years

Love-hate relationship with Barcelona. PROs: experiencing a mix of urban grunge and rich Spanish history was refreshing; plus @FAccTConference was lovely CONs: my laptop bag (and its contents) found a new home, thanks to ladrones. I hope the new owners enjoy my (ex-)machine 😭

1

0

7

@umangsbhatt

Umang Bhatt

3 years

There is a bunch of great existing experimental and theoretical work in this direction and even a few discipline-specific review papers. Our current review, while likely imperfect, helped us organize our thoughts, and we hope others find it helpful too!

1

1

7

@umangsbhatt

Umang Bhatt

3 years

@JessicaHullman By any chance, do you know of work that explicitly considers separately visualizing epistemic and aleatoric uncertainty? @javiac7 made the attached for however, we didn't come across user studies that consider displaying these two types of uncertainty

Tweet media one

2

0

7

@umangsbhatt

Umang Bhatt

3 years

It’s always fun collaborating with folks from tons of organizations: @CambridgeMLG @LeverhulmeCFI @PartnershipAI @IBMResearch @CarnegieMellon @Element_AI @IntelAI @nyutandon @turinginst @SonyAI_global

1

2

7

@umangsbhatt

Umang Bhatt

3 years

In the paper, we start by explaining the different types of uncertainty that can arise when working with ML models. We discuss how to estimate and quantify these uncertainties and how to evaluate the quality of our estimates.

Tweet media one

1

1

7

@umangsbhatt

Umang Bhatt

3 years

After moving to the Cathedral of Bayes ( @CambridgeMLG ), I realized that I grossly undervalued the importance of uncertainty. Simultaneously in work with @PartnershipAI on explainability, we had multiple stakeholders mention their desire for well-calibrated uncertainty estimates.

1

1

7

@umangsbhatt

Umang Bhatt

5 years

It’s nerve-racking to see people from different parts of your life meet: you’re unsure if their interactions will be like napalm or like fireworks. But with proper priming and timing, the latter can be so heartwarming to see

0

0

6

@umangsbhatt

Umang Bhatt

3 years

We then discuss leveraging uncertainty for (1) fairness specifically measurement and sampling bias, (2) decision making with special care for mode failures, and (3) trust formation in human-AI teams.

1

1

6

@umangsbhatt

Umang Bhatt

3 years

We close with common pitfalls when communicating uncertainty ( @d_spiegel has seminal work on this) and common visualization methods.

1

1

6

@umangsbhatt

Umang Bhatt

3 years

Love @krvarshney 's framing that uncertainty can be "intellectual humility" for AI

@krvarshney

Kush Varshney कुश वार्ष्णेय

3 years

Uncertainty quantification adds much-needed intellectual humility to AI. Visit the @AIESConf poster "Uncertainty as a Form of Transparency: Measuring, Communicating and Using Uncertainty" by @QVeraLiao @prasatti @umangsbhatt and others today to learn more.

1

7

46

0

2

6

@umangsbhatt

Umang Bhatt

5 years

I'm also at #NeurIPS2019 @NeurIPSConf in Vancouver this week if you want to discuss explainability, fairness, and human-AI teams. Feel free to reach out! I'll be presenting a subset of this as a poster at the #HCML2019 workshop on Friday (Dec. 13) in West Level 2, Room 223-224

2

1

6

@umangsbhatt

Umang Bhatt

3 years

I hope uncertainty becomes a keystone in the FAccT community: specifically, uncertainty ought to be as important as explainability in the world of ML transparency @trustworthy_ml #XAI

1

1

6

@umangsbhatt

Umang Bhatt

3 years

We will also have a panel on the implications of the new @EU_Commission AI regulation

1

1

6

@umangsbhatt

Umang Bhatt

5 years

@Aaroth We ran a survey around how people use different explainability techniques in practice. Feature importance literature seems most “mature” — explanations aren’t exposed to end users but are used by ML engineers to sanity check. Peep

0

1

6

@umangsbhatt

Umang Bhatt

6 years

Super excited to be at #SOCML2018 at Google Toronto - I’d love to talk to you if you’re here and learn from you! Thanks for organizing it @goodfellow_ian :)

0

0

6

@umangsbhatt

Umang Bhatt

6 years

When life sacks you and sets you back 16 yards, get right back up and throw a 41-yard touchdown just like @Tuaamann_ #NationalChampionship

0

0

2

@umangsbhatt

Umang Bhatt

2 years

@adrian_weller @Cambridge_Uni @turinginst @CDEIUK @LeverhulmeCFI @CambridgeMLG @Cambridge_Eng @Cambridge_CL @Marion_InfoLaw Congratulations Adrian!!! 🥳 What a well deserved honor 🚀

0

0

6

@umangsbhatt

Umang Bhatt

3 years

So last June, we set out to wrap our heads around the uncertainty literature from various disciplines: Bayesian Stats (Prior Lovers 😊), Frequentist Stats (Glorified Counters 😳), Viz, HCI, etc. With folks from each area, we put together a review, but we struggled… a lot.

1

1

6

@umangsbhatt

Umang Bhatt

5 years

@geomblog We ran a survey at the @PartnershipAI around this (joint with @fiddlerlabs ). Check out Section 4.1.2:

Tweet card media

Explainable Machine Learning in Deployment

Explainable machine learning offers the potential to provide stakeholders with insights into model behavior by using various methods such as feature importance scores, counterfactual explanations,...

1

1

6

@umangsbhatt

Umang Bhatt

5 years

I thought texting and driving was unsafe until I just saw a man trying to write on an index card with a sharpie while driving 🤯

0

0

6

@umangsbhatt

Umang Bhatt

3 years

This virtual event is co-organized with Plamen Angelov @autreche @hima_lakkaraju @nuriaoliver @stratis_ @adrian_weller and co-sponsored by @ELLISforEurope @turinginst @LeverhulmeCFI

1

2

6

@umangsbhatt

Umang Bhatt

3 years

Might be of interest @trustworthy_ml @black_in_ai @_LXAI @AiDisability

0

1

5

@umangsbhatt

Umang Bhatt

2 years

@dan_w_ley @Harvard @hseas @hima_lakkaraju @adrian_weller @DanMagazzeni @lenachossss Big congrats to @dan_w_ley , @hima_lakkaraju , and @ai4life_harvard . Can’t wait to see what you do in the other Cambridge!

0

0

5

@umangsbhatt

Umang Bhatt

6 years

"Predictive risk assessments are a new form of algorithmic violence" - what a powerful and salient call to arms for the use of causal inference #FAT2018

0

3

4

@umangsbhatt

Umang Bhatt

4 years

So well deserved — Alice is an incredibly humble, brilliant researcher and a thoughtful, supportive mentor. Grateful to be able to work with her!

@alicexiang

Alice Xiang

4 years

Honored to be featured in this special report from Trix Magazine on AI ethics and inclusion:

2

7

25

0

0

5

@umangsbhatt

Umang Bhatt

3 years

This panel will feature @vrbenjamins ( @Telefonica_En ), @emiliagogu ( @EU_ScienceHub ), @boom_lawyered ( @Clifford_Chance ), Gabriele Mazzini ( @EU_Commission ), and @mikarv ( @UCLLaws ): should be a delight! Register to attend at

Tweet card media

ELLIS HCML 2021

A day-long workshop of talks and panels on human-centric machine learning. The workshop took place virtually. We thank all speakers, panelists and attendees for the participation! The recording of...

sites.google.com

@umangsbhatt

Umang Bhatt

3 years

We will also have a panel on the implications of the new @EU_Commission AI regulation

1

1

6

0

0

5

@umangsbhatt

Umang Bhatt

6 years

First #steepedmoment on a cool evening in Pittsburgh. Super easy to make and definitely beats all other to-go single serve coffees I've had. Can't wait to use it when I travel. Thanks @steepedcoffee

0

1

4

@umangsbhatt

Umang Bhatt

3 years

@peterbhase @__Owen___ Awesome resource! You might be interested in our ICLR 2021 paper on explaining uncertainty estimates or our IJCAI 2020 paper on evaluating feature importance explanations

Tweet card media

Evaluating and Aggregating Feature-based Model Explanations

A feature-based model explanation denotes how much each input feature contributes to a model's output for a given data point. As the number of proposed explanation functions grows, we lack...

1

1

5

@umangsbhatt

Umang Bhatt

5 years

Excited to be visiting @adrian_weller and the rest of the @Cambridge_Uni MLG later this week. Will be giving a short talk () on Thursday on some recent work with José and Pradeep regarding how to aggregate local feature-level explanations.

1

0

5

@umangsbhatt

Umang Bhatt

5 years

🤦🏾‍♂️ Find and replace all instances of “we” with “I” — repulsive attempt at science

@AndrewM_Webb

Andrew M. Webb

5 years

So in @sirajraval 's livestream yesterday he mentioned his 'recent neural qubit paper'. I've found that huge chunks of it are plagiarised from a paper by Nathan Killoran, Seth Lloyd, and co-authors. E.g., in the attached images, red is Siraj, green is original

Tweet media one

Tweet media two

193

1K

4K

0

0

5

@umangsbhatt

Umang Bhatt

4 years

Huge shoutout to @melechuga @baricks @adrian_weller for helping make this happen

0

0

5

@umangsbhatt

Umang Bhatt

3 years

We will hear talks from @BullardKalesha @Chapfuwa @mrtz @erichorvitz @ecekamar @fjmarmole @ChinasaTOkolo Adish Singla @ChenhaoTan @viegasf

1

1

5

@umangsbhatt

Umang Bhatt

3 years

Oh, and this work will appear at @AIESConf in May #AIES2021

0

1

5