Chhavi Yadav
@chhaviyadav_
Followers
2,118
Following
2,835
Media
46
Statuses
1,866
Machine Learning Researcher | PhD Candidate @ucsd_cse | @trustworthy_ml
Joined July 2015
Don't wanna be here?
Send us removal request.
Explore trending content on Musk Viewer
Afghanistan
• 506683 Tweets
Arlington
• 337731 Tweets
Tulsi
• 263443 Tweets
台風の影響
• 245233 Tweets
#WWERaw
• 51794 Tweets
LU ESTAMOS CON VOS
• 46934 Tweets
Zuck
• 34913 Tweets
#FranColapintoaF1
• 17331 Tweets
O Vasco
• 17040 Tweets
ニンダイ
• 16870 Tweets
Capitán
• 16670 Tweets
Milton
• 14684 Tweets
Pinned Tweet
Delighted to receive a Best Paper Award 🏆for my latest work — FairProof : Confidential and Certifiable Fairness for Neural Networks ()
— at the Privacy-ILR Workshop
@iclr_conf
! (my 1st :)
Will also be presented
@icmlconf
Slides :
14
14
195
Opened my Google scholar to a pleasant surprise today - my first paper is now 100 citations old! 😀
2
3
195
Excited to start my PhD
@ucsd_cse
as a Jacobs Engg. Fellow! 👩💻🔍
Looking forward to making ML more robust & safe with
@kamalikac
(& beaches post-covid 🏖️)
Thanks to the never-ending love & support from mentors, family & friends 🙏
21
0
177
Officially became a PhD candidate today!! Really excited for what's to come in the future :) 🎉🎊
14
1
170
The dates for neurips coincide with semester end exams & application deadlines. It's really frustrating to not be able to completely enjoy the conference & adds so much pressure. Last year I was applying for PhD programs & this year I am writing exams 😑
@NeurIPSConf
11
3
141
1/2 My first ever paper is out on arxiv finally! For all the people dealing/have dealt with MNIST, here's an extra never seen before 50,000 digits that you can test your models on! Check out:
4
21
136
If you are tired of tweets on sentience, here is something more concrete to read - my latest paper ‘A Learning-Theoretic Framework for Certified Auditing of Machine Learning Models’ A Thread 1/n
2
28
128
Did you know QMNIST is now a part of Pytorch Torchvision.Datasets? You can use it as easily as MNIST, has 6x the test set and also comes with meta data like writer id!! 😀
6
14
105
Super excited for our
@WiMLworkshop
Unworkshop session on "Explainable machine learning: do we have the right tools?" with
@ML_Theorist
Come join us for an interesting discussion with
@hima_lakkaraju
@james_y_zou
@sameer_
@CynthiaRudin
Timing - this wednesday at 12:25-1:30 pm ET
1
22
98
Delighted to announce that our paper 'Cold Case: The Lost MNIST Digits' has been accepted as a Spotlight at
@NeurIPSConf
. Don't forget to test your systems on QMNIST!!
Paper :
Code & Dataset :
7
13
91
1/ Looking for all the twitter discussion on Bayesian DL? I compiled them for ya
New blog post with
@carlesgelada
-- "A Sober Look at Bayesian Neural Networks":
Without a good prior, Bayesian uncertainties are meaningless. We argue that BNN priors are likely quite poor, and concretely characterize one specific failure mode.
10
93
391
2
15
89
I have a bayesian joke but it's updating..
I have a deep learning joke but it has a lot of layers to it.
27
51
492
0
1
85
Decision season thoughts : How many of the ML professors of today would get in if they applied to grad schools in 2019?
4
1
80
Delighted to have won the ‘Excellence in Contributions to Diversity’ award
@ucsd_cse
! Thanks so much for recognizing my efforts towards enhancing the diversity in our department.. This is certainly a morale booster & a positive sign to continue the efforts 🙂
13
1
78
Announcing our Neurips 2023 workshop "XAI in Action: Past, Present, and Future Applications"!
Apart from the usual submissions, we have a special "demo" track to demo s/w libraries, visualization tools, applications of ML w major XAI component
Ddl :22 Sept
You are back from ICML, submitted your rebuttal to NeurIPS and wondering what to do next? Submit your work to the NeurIPS AXAI workshop
1
5
30
3
22
75
India is suffering from an apocalypse & it needs help💔. Here are some links to initiatives that we can support :
1.
2.
3.
4.
3
29
71
Research Internship opportunity in India for undergrad/master's students with some of the best advisors one can think of!
Do apply -
2
20
64
This whole reviewing thing has become a total waste of time, both as an author (reviews are random & non-sensical) & as a reviewer (papers in my lot were extremely bad). We as a community need to figure out something, atleast stop using "publications" as a metric
#icml
😩
4
1
64
@JuliaZeitlinger
@nas_pari
@AnimaAnandkumar
omg r u kidding me, wth is wrong with this picture? Proves once again, whether bikini or dressed from head-to-toe, it doesn't really matter, sick minds keep coming up wid sick norms abt wat others shud do. Ain't no1 needs to heed to this shit. plz stand against, not perpetuate bs
0
1
62
@AnimaAnandkumar
Thank you so much for sharing. this gives all of us, but specially young Indian women PhDs, the courage to fight back. Thanks a lot ❤️
0
1
51
Some reviewers are so nice. They start with the positives in the paper & tell you how you can 'improve' your paper and make your claims more believable. They do so much thinking for you. This is the reviewer I look upto. They restore your faith in humanity ❣️
1
2
51
The 14th **Asian Conference on Machine Learning** is coming to India! It will be in a hybrid mode, offline meeting in Hyderabad.
Submit your works here!
There are both journal (ddl 26th May)and conference (ddl 23 June) tracks..
Check for more details!
1
8
51
Excited to organize my first workshop
@NeurIPSConf
. This time around the workshop will focus more on surveys and analyses. CFP soon. Checkout this space
Excited to share the workshop proposal put together by
@MayooreJ
@chhaviyadav_
@PrabhuPradhan
@JesseDodge
me and a few others (waiting for their
@Twitter
handles to tag them) on
@MLRetrospective
has been accepted to
@NeurIPSConf
#NeurIPS2020
!! Stay tuned for the CFP details!
1
2
26
1
7
50
Hi friends! I'll be
@NeurIPSConf
from 11th - 16th!
Would love to meet people and chat about all things Trustworthy AI including XAI, AI Safety & Security. HMU if interested! :)
Also come to the XAI workshop I'm organizing on Saturday
#NeurIPS2023
#NeurIPS
3
5
48
Join us 2mrw @⚡️XAI-in-Action : Past, Present & Future Applications⚡️ workshop
@NeurIPSConf
for an exciting discussion on how far we've come, the current state & future of XAI !
Bonus: Live Demos!
Submit ur ❤️🔥 panel questions:
🔗
1
11
44
Please consider submitting to & attending our ML-RSA workshop
@NeurIPSConf
.
Deadline - 23 Oct. We invite papers from all subfields of ML.
Amazing speakers include
@rzshokri
@adjiboussodieng
@mariadearteaga
@sina_lana
@ShibaniSan
@shakir_za
#KilianWeinberger
& more!
🎉Our
#NeurIPS2020
workshop website is UP!! 🎉
This time we're focusing on analyses/surveys (but still want more retros)! We hope both unpublished and seasoned researchers submit to the workshop to help synthesize knowledge in this fast-paced field.
👇
1
15
43
0
13
45
This was a lot of fun! So honored to be the host for the first talk & officially kickstart the amazing series! Recorded talk is available on youtube here :
1/ Attended Percy Liang’s (
@percyliang
) talk today and want to continue the discussion on robust ML? Looking to meet new people working in this field? Use this thread to discuss what you thought of the talk or any questions you have, and make some new connections 👇🏾
1
15
61
2
4
43
I am organizing a workshop
@NeurIPSConf
- ! We have received amazing papers in a *wide* variety of areas in ML. If you are interested in being an expert reviewer, DM me.
2
5
43
In CS we don't say "I love you", we send relevant papers & help debug code & I think that's beautiful 😌💕
In physics we don't say "I love you" we say "I think your solution makes sense” and I think that’s beautiful
66
284
3K
0
0
38
We are looking for reviewers for our
#XAI
-in-Action workshop at
@NeurIPSConf
2023!
If you promise to be a good reviewer (read thoroughly & give genuine reviews etc.) please fill out this form - !
1
11
37
Very excited for this session! We are joined by the amazing
@rosemary_ke
@FrancescoLocat8
@k__niki
@anirudhg9119
& Stefan Bauer as well! Please come to discuss Disentangled Representations with us all
@icmlconf
@WiMLworkshop
#ICML2020
attendees, join us for our virtual breakout session
@wimlworkshop
on "Entangled Conversations on Disentangled Representations (EnCoDR)"
When: July 13 (Mon) @ 9:45-10:45AM PST/4:45-5:45PM GMT
w/
@chhaviyadav_
@laure_delisle
Irina Higgins & Jovana Mitrovic
0
2
14
0
9
37
Super excited to be a part of this amazing effort. I will be managing the bi-weekly seminar series (), soon to launch in mid October. If you have ideas on how we can encourage active discussion between participants, send them my way!
📢 Excited to announce our launch as the Trustworthy ML initiative (TrustML) . Our efforts just got bigger and better! Read on 👇[1/n]
#trustworthyML
#MachineLearning
#ArtificialIntelligence
#DeepLearning
5
75
113
0
3
35
@iclr_conf
is going sooo smooth. I really wish we have both virtual & actual conference versions in the future. This is an amazing experience! Thanks
@kchonyc
@shakir_za
@white_martha
@dawnsongtweets
@srush_nlp
& all the other organizers!
Please join me in expressing admiration and thanks to the
#iclr2020
orgs for their amazing pivot to a virtual conference. As a fellow 2020 chair, this was a monumental feat in minimal time. Thank you
@srush_nlp
@shakir_za
@dawnsongtweets
@kchonyc
@white_martha
and all other orgs!
4
47
450
0
5
33
So u know of Occam's Razor.. Ever heard of Hanlon's Razor though?
"Hanlon’s Razor states that we should not attribute to malice that which is more easily explained by stupidity. In a complex world, this helps us avoid paranoia and ideology. 1/3
2
3
33
2
0
31
Can't thank enough all the amazing people in my life who came to support me during my talk at
#NeurIPS2019
. Thanks a lot to
@ShengLiu_
@backpropper
@GloriaMeltemi
@activatedgeek
@_prieuredesion
❤️🐥❤️🐥❤️🐥
1
0
30
I'll be hosting too! Please join. looking forward to it!
Reminder to
@NeurIPSConf
attendees: our Social is scheduled just about 24 hours from now: 3-5pm ET tomorrow! Tune in for some exciting virtual activities and discussion. Joining info in the main thread below👇🏽
0
8
37
0
2
30
@thegautamkamath
I recently checked the recipients of a fellowship that i applied to. All the awards went to students with famous advisors & with fancy university tags. I wonder if anyone is actually looking at the work people are doing or just the tags.
1
0
28
Can we stop being so ridiculous & have some empathy? The most insensitive blurb I've read in a while. Whosoever wrote this down, has no idea what me & more people are going through. Thanks
@hima_lakkaraju
for bringing this up
As I struggled to deal with the impact of COVID on my family members in India, I got delayed by a day for submitting my reviews for a conference & I got a message from a senior reviewer with the blurb below. My humble request to everyone - pls don't say this to anyone ever! [1/n]
1K
6K
39K
0
1
28
how do people manage to grow their technical expertise outside of their current research? give your best tips.. I almost can't find time
#phdchat
4
0
28
🍵🍿Ahem true story : someone I know received an
#ICML
review which is a **ditto** copy of their
#ICLR
review, that is publicly available ofc. And the authors obviously realized this! 👏🏻 Guess who's in trouble now ?? 🤔
They say you need brains to cheat as well 😂
This whole reviewing thing has become a total waste of time, both as an author (reviews are random & non-sensical) & as a reviewer (papers in my lot were extremely bad). We as a community need to figure out something, atleast stop using "publications" as a metric
#icml
😩
4
1
64
5
0
26
If anyone is interested in co-leading a WiML Un-Workshop session with me
@icmlconf
, please reach out.. Here is the call :
@WiMLworkshop
4
9
27
@keenanisalive
It's also really sad to see professors turn down great applicants for those from the top schools without interviewing either.
1
0
25
Want to know about verification of fairness of a model while maintaining model confidentiality? Come to our poster tomorrow, 24 July in Hall C from 11:30-1pm
#2304
#ICML
@icmlconf
@WiMLworkshop
0
5
25
Excited for the post-COVID (hopefully we're there now) travel!
I am in NY for my summer internship.. would love to meet some people working on ML and aligned areas!
Who should I meet? Also hit me up if you would like to meet me~
2
1
24
Join us tomorrow @ our
@WiMLworkshop
breakout session titled
'Entangled Conversations on Disentangled Representations'
If u have any questions u would like to discuss, feel free to leave them below!
July 13 (Mon) @ 9:45-10:45AM PST/4:45-5:45PM GMT
Link:
@icmlconf
virtual site
We are excited to introduce the topics for the third breakout session at the
#WiML2020
Un-Workshop co-located with
#ICML2020
, which will take place on Monday, July 13 at 4:45PM – 5:45PM GMT. More information here:
#WiML
#machinelearning
@icmlconf
0
9
21
1
7
24
Excited to have become one of the 'top reviewers' at the first ever proper conference I reviewed for. But the irony is most of my reviewers were really bad~ 🤦♀️
2
0
21
Did you register yet? It's gonna be super exciting I swear.. I will be leading a breakout room on explainability + interpretability!
Looking forward to it!
1/ 📢 Oct 28 is a very special day! It's The Trustworthy ML Initiative's one-year anniversary 🥳🎂
To celebrate, we're holding a special event jointly with Montreal AI Ethics Institute (
@mtlaiethics
): The Invisible Elephant 🐘 in the Room: The Trustworthy ML Un-Symposium 🎉
1
22
68
0
5
21
As I struggled to deal with the impact of COVID on my family members in India, I got delayed by a day for submitting my reviews for a conference & I got a message from a senior reviewer with the blurb below. My humble request to everyone - pls don't say this to anyone ever! [1/n]
1K
6K
39K
4
0
19
Every time I’m told that I have an ‘accent’, can’t help but think - u too have an accent, evry1 has an accent, all accents r equal none is superior, ppl from other countries might have a problem understanding your accent too!
Usually I don’t say this aloud but i think i need to..
4
0
20
I’m going to get harassed for saying this but every woman should know how to love herself unconditionally first, have a career and fight for her rights. It’s part of being an incredible nurturer and caregiver.
I’m going to get harassed for saying this but every woman should know how to do a combination of bench science, field work, and bioinformatics. It’s part of being an incredible nurturer and caregiver.
3
36
358
0
2
19
First talk of the day by
@sameer_
has begun!!
Join us at the XAI-in-Action workshop in Room 271 !
1
2
19
Will be attending
@icmlconf
#ICML
🏖️🌴! Would love to meet people and discuss anything
#trustworthyAI
:) Please DM me to meet :)
0
1
19
Join me today for an interesting discussion on 'Challenges & Opportunities in Certified Auditing' at 11am, Exhibit Hall G Level 100 Exhibition Halls
@WiMLworkshop
0
4
19
If u'r attending ICML in-person, come meet me at
@WiMLworkshop
!
- I'll be leading an in-person breakout session on 'Challenges & Opportunities in Certified Auditing' at 11:05 am, 18th Monday, Hall G level 100
- Presenting a poster 10:10-10:40am & 4:40-5pm EST 18th Monday on 1/2
The 3rd WiML UnWorkshop at ICML is just a few days away! All of this is possible thanks to our sponsors
@Apple
@DeepMindAI
@DEShawResearch
@HomeDepot
@Google
@Meta
@MSFTResearch
and
@quantumblack
Schedule:
#WiML2022
#ICML2022
0
5
15
1
2
16
yay! so excited to have been a part of this initiative! 🎉🎊
Yayy! Can't believe it's been a year already. Super grateful to be working with this amazing team
@trustworthy_ml
@JaydeepBorkar
@sbmisi
#SarahTan
@sarahookr
@_cagarwal
@chhaviyadav_
@m_lemanczyk
1
4
40
1
1
16
Had quite a blast at the discussion! Thanks to our guests
@hima_lakkaraju
@sameer_
@james_y_zou
@CynthiaRudin
+ dear attendees for making it super fun :)
Watch this space for a summary, we'll post soon!
Super excited for our
@WiMLworkshop
Unworkshop session on "Explainable machine learning: do we have the right tools?" with
@ML_Theorist
Come join us for an interesting discussion with
@hima_lakkaraju
@james_y_zou
@sameer_
@CynthiaRudin
Timing - this wednesday at 12:25-1:30 pm ET
1
22
98
2
2
16
@animesh_garg
@yisongyue
Classic case of - "When a measure becomes a target, it ceases to be a good measure"
0
0
15
Good News for those submitting to the XAI-in-Action
@XAI_in_Action
Workshop
@NeurIPSConf
!
We have extended our submission deadline to **2nd Oct** (instead of 22 Sept previously).
Please help us spread the word!
Announcing our Neurips 2023 workshop "XAI in Action: Past, Present, and Future Applications"!
Apart from the usual submissions, we have a special "demo" track to demo s/w libraries, visualization tools, applications of ML w major XAI component
Ddl :22 Sept
3
22
75
1
9
15
Thanks to all the people who attended our social and made it super fun! 🏖️🎊🌴
Amazing turnout for our sunset beach social at ICML 🏝️🌺🍹☀️So great to see a community come together. And someone even brought a poster!!!
1
5
62
0
0
15
Sorry
@NeurIPSConf
but the best mug award goes to
@WiMLworkshop
without a doubt!
Looking forward to having my tea in this 🍵
0
2
14
'Publish or Perish' IS perishing good science. It is the reason why many smart people do NOT want to pursue/continue research. & at this rate of publication, peer review is hanging by a string..
0
2
14
Murphy's law - some random cuda memory issue will show up for previously bug-free code when you are in a time crunch.
1
0
14
Has reviewing always been adversarial?
-> Lately it seems like the rating is harsher than the review itself.
-> Also just because your list of 'weaknesses' is long, doesn't mean the paper has that many weaknesses.. shouldn't reviewers adjust for the possibility of being wrong 1/2
1
0
13
@sarameghanbeery
@AnimaAnandkumar
@zoom_us
As an organizer previously, I thought about possible ways of zoombombing. TBH there are many & it doesn't even take a smart person.
@zoom_us
you need to make ur platform more secure if u don't want to become another which unsafe for minorities. U can't leave this to goodwill
2
2
13
Code and the reconstructed dataset can be found at :
0
1
13
@chrisalbon
In my home country students are taking classes on mobile phones with slow internet. In a household with daughters & sons, sons are being given preference to access classes & girls are left to do household chores. Cannot fathom how parents can do this to their own daughters
0
1
12
Our last invited talk for the day from
@leilanigilpin
on XAI for Autonomous Vehicles!! 🚗
Join us in Room 271 to participate!
1
0
11
If Tokyo Drift is not played on my funeral, I am not leaving. Does anyone else feel the same about Tokyo Drift ????? 😍 I probably want to visit Tokyo Just bcoz of this 😍🏎🚈🛒
2
0
12
0
0
12
Some great coding rules & conventions listed at (Thanks to
@activatedgeek
for pointing out)
My top 20 below, Enjoy!
1. The essentials of programming: No surprises, minimize coupling & maximize cohesion
2. Don't confuse ease of learning with ease of use
1
1
11
We're indeed v delighted & honored that
@AmandalynneP
,
@rajiinio
,
@emilymbender
,
@cephaloponderer
&
@alexhanna
chose our workshop to submit their excellent piece on datasets! Highly recommend to check this out. The timing couldn't have been any better.
Award for forwarding the discussion around collecting and using data:
"Data and its (dis)contents: A survey of dataset development and use in machine learning research"
Amandalynne P., Deborah Raji, Emily M. Bender, Emily Denton, Alex Hanna.
2/4
1
24
90
1
2
11
With this India becomes the first country to land on the lunar poles of the moon and the forth country to ever to do a controlled landing on the moon! 🇮🇳
@isro
thank you 😊
Chandrayaan-3 Mission:
'India🇮🇳,
I reached my destination
and you too!'
: Chandrayaan-3
Chandrayaan-3 has successfully
soft-landed on the moon 🌖!.
Congratulations, India🇮🇳!
#Chandrayaan_3
#Ch3
70K
284K
850K
0
0
10
Had a very fun time moderating the amazing
@marinkazitnik
Thank you so much
@marinkazitnik
for sharing your research and journey at today’s
#TrustML
seminar 🎉🎉
Recording of today’s talk “Graph Representation Learning for Biomedical Discovery” is now available
1
1
8
0
1
11
super excited to moderate this talk!
In line with ICML's ML for science theme , another amazing talk to add to your collection :)
📢 1/ We are delighted to have Marinka Zitnik
@marinkazitnik
from
@Harvard
with us for the next TrustML seminar on Thursday, July 22nd, 12pm ET! 🎉🎉
Register here: .
Check this thread for the speaker and talk details! 👇
1
4
22
0
3
11
The relation between RL (Reinforcement Learning) and causal inference has been a topic of some debate. It can be resolved, I believe, by understanding the limits of each. RL authors call this limit "off-policy". I like "off-knobs", imagining a machine with finite number of knobs.
4
66
238
1
0
10
@j_foerst
I would be really interested in academic offers. This is perhaps way more closed information that atleast I have no sense of. Also year-wise separation of offers & emerging trends will be a good addition to the dashboard.
@deviparikh
@abhshkdz
1
0
10
@AnimaAnandkumar
@MeTooSTEM
@Caltech
I really hope
@Caltech
takes a stand.
@AnimaAnandkumar
I read your post , you are an inspiration for the girl squad it takes so much courage to do what you have been doing for a long time & continue to do
0
0
10
@AnimaAnandkumar
@TelleWhitney
@suchisaria
The parties are only getting bigger and crazier . Attaching some proof 😉
Highlights for me :
- Met some amazing women who left me inspired
- Presented my first poster
- Talked to
@JeffDean
about ML
- Watched Spot live in action
@BostonDynamics
0
1
10
@AnimaAnandkumar
@zoom_us
Gosh disgusting. Ppl need to get a life & stop stooping to such levels. I never imagined this could happen within academia. Must be really terrible people with a sick mind who thought of doing this. totally unacceptable behavior
1
1
10
Booked tickets for my mother yesterday from
@airindiain
& I am already so disappointed. Want to make a minor edit to the passenger name but apparently their system doesn't allow them to do it. Now the only way out is to cancel the ticket, bear the hefty refund fee & rebook..
4
0
10
Excited to be giving this talk
@trustworthy_ml
!
Come & celebrate our 2nd anniversary event with us! Register here -
Check the full program here -
We are excited to highlight
@chhaviyadav_
who will be giving a talk as a part of our "Celebrating Young Researchers" event on our 2nd anniversary this coming Thursday, October 27! 🎉
Register to get Zoom link:
0
2
7
0
2
9
Happening today! Poster at Hall C, Wed 11:30 session
#ICML2024
Machine learning models must be audited for fairness, robustness and accuracy, yet many are confidential. How can we audit a confidential model for fairness while still retaining confidentiality? We show how to do this in our
#ICML2024
paper.
1
6
46
0
1
9
So I asked ChatGPT to write a poem similar to twinkle twinkle little star.. and the result is beautiful!
0
1
9