![Erin Grant Profile](https://pbs.twimg.com/profile_images/1534095330580566016/bIHk0KQ8.jpg)
Erin Grant
@ermgrant
Followers
3K
Following
7K
Media
19
Statuses
279
🚨 I'm no longer active on Twitter. Find me on 🦋: @[email protected] Senior Research Fellow @GatsbyUCL & @SWC_Neuro {learning, representations} in 🧠💭🤖
London, England
Joined July 2015
I gave my dissertation talk last month! 🥳 Thanks to my advisor Tom (@cocosci_lab), my committee, and the communities at Berkeley, Princeton, and beyond for fun and support along the way ❤️
37
8
400
Postponing the #NeurIPS2020 deadline is up to the PCs, but our community at large can still take action to help. If you intend to send your work to @NeurIPSConf & are #BlackInSTEM, I can try to help in the final period before the deadline (mock review; copy-edit)--DM me or email.
3/ I fear that if we don't do so, our coming NIPS will be much less diverse, dominated by authors who have been least affected by the disruptions. One way we can support those in need, possibly with the smallest affect on the tight NIPS timeline, would be to offer extensions….
3
31
162
#ICML2021, @thirdreviewer: Inductive biases do not affect generalization behavior, only data does ie. all models are the same. Reject w/ max confidence.
11
3
120
Introducing "visual selectivity" as a point of behavioral comparison between humans & neural networks, in this "real tour-de-force" (#NeurIPS2021 meta-reviewer) led by @ThomasALanglois* & Charles Zhao*! w Ishita Dasgupta, Tom Griffiths (@cocosci_lab), @norijacoby (@CompAudition).
We are thrilled to share our upcoming oral in NeurIPS 2021! Passive Attention in Artificial Neural Networks Predicts Human Visual Selectivity by @ThomasALanglois*, Charles Zhao*, @ermgrant, Ishita Dasgupta, @cocosci_lab & @norijacoby @CompAudition @MPI_ae. (1/10)
1
18
100
Super excited that I will be postdoc'ing starting Fall'22 as a senior research fellow at @SWC_Neuro + @GatsbyUCL with @SaxeLab 🥳🇬🇧 Please send along relocation / neighbourhood / coffeeshop advice for London! 🙂.
14
2
100
The 5th edition of the Meta-Learning Workshop (#MetaLearn2021) is taking place on Workshop Monday (13th Dec.) @ #NeurIPS2021, and the CfP is now out! Please submit your up-to-8-page research papers by Sept. 17th; more details at
1
20
95
We're running an Equal Access to Application Assistance (EAAA) program for this fall's application cycle to higher-degree programs (MS/PhD) at @Berkeley_EECS!. Any applicant to @Berkeley_EECS can submit their application packet for review by Nov. 1st at
2
22
89
Here we go! 🧠💭🤖 Re-Align starting in Lehar 2 @iclr_conf now 🤩
Really looking forward to tomorrow's First Workshop on Representational Alignment (Re-Align) at #ICLR2024! . Check out the schedule at w/ @ermgrant @sucholutsky @lukas_mut @achterbrain.
1
7
84
I support pathways to research in AI for young people. However, to echo others: I think the likely outcome of this no-feedback high school submission track will be to give participation prizes for well-resourced high school students. To add on:.
This year, we invite high school students to submit research papers on the topic of machine learning for social impact! See our call for high school research project submissions below.
1
9
80
Looking forward to the #NeurIPS2018 Workshop on Meta-learning! It's expected to be the largest on Sat (>600 attendees). Speaker and poster details now online at Please add/upvote a panel question! w/ Frank Hutter, @sachin_ravi_, @joavanschoren, @janexwang.
2
15
70
Though we've heard many stories (visa refusals, long wait times), we don't yet have data on how many attendees face difficulties attending @NeurIPSConf in person. Please help us advocate for reducing barriers to participation by filling out our survey: ➡️.
We're launching a survey on barriers to in-person participation at NeurIPS to inform potential solutions to the difficulties of prospective in-person attendees. If you faced visa or other difficulties attending #NeurIPS2023, please complete the survey at:
2
17
52
Pre-doc students can join Gatsby for a summer to learn about the mathematical tools useful for ML & neuro ⬇️ I had a lot of fun co-teaching part of this course with @ClementineDomi6 last year, and am looking forward to this year's cohort!!.
📢 Apply now for the 7-week #Mathematics #SummerSchool!.You will develop mathematics skills/intuition necessary to enter the #TheoreticalNeuroscience or #MachineLearning field & have the opportunity to interact with our researchers. Where: @SWC_Neuro.ℹ️
0
6
31
@neu_rips @NeurIPSConf As the artefacts presented at machine learning conferences (papers, tutorials, etc.) are increasingly about closed-source product research, and the conferences become closer to trade shows, an industry majority in leadership appears less surprising.
1
1
25
The #NeurIPS2023 barriers-to-participation survey is still open! Please fill it out, especially if you are unable to join us in NOLA due to a visa barrier. I'll present some preliminary stats on this (& and on finaid) at the Town Hall tomorrow (.
We're launching a survey on barriers to in-person participation at NeurIPS to inform potential solutions to the difficulties of prospective in-person attendees. If you faced visa or other difficulties attending #NeurIPS2023, please complete the survey at:
0
10
25
The #NeurIPS2020 deadline has been extended by 48h. If you can use this time to help out your @black_in_ai colleagues, please add your name to this spreadsheet (thanks @le_roux_nicolas!):.
1
8
25
@WiMLWorkshop @ #NeurIPS2020 will be virtual this year!.Poster presenters & volunteers will receive registration for the whole #NeurIPS2020 conference, so please don't hesitate to submit your recent work / sign up to volunteer!.
Call for Participation! The 15th WiML Workshop will be co-located with the virtual @NeurIPSConf in December, 2020. Deadline for submission: September 28th, 2020.Read the full CFP: #WiML #WiML2020 #NeurIPS2020
1
7
25
The deadline for extended abstract submissions to our @iclr2019 workshop is tomorrow, midnight AoE! Looking forward to discussing modularity, structure, abstraction & transfer in RL with researchers from diverse backgrounds incl. machine learning, neuroscience & cognitive science.
CfP for the @iclr2019 workshop on structure and priors in reinforcement learning (SPiRL), deadline 3/7!
0
5
21
🌟 Calling all mentees AND mentors for #WiML's pilot mentorship program! 🌟 We're turning the very popular mentorship roundtables at @WiMLworkshop into something more long-term.
📣 We are excited to announce the ***Spring 2022 #WiML Mentorship Program***! If you are interested in participating in the program either as a mentor or as a mentee, please sign-up using this form: Deadline to sign-up: February 1st.
1
4
22
@AndrewLampinen moderates the last panel with @aran_nayebi @mtoneva1 Klaus Greff @s_y_chung on bridging / increasing: What drives neural predictivity? Can we (should we) use understanding of neural predictivity to make our models more biologically plausible (for science or eng)?
1
2
20
Deadline is tonight, 11:59 PM anywhere on Earth! We look forward to your submissions!.
Excited to put out a call for the #NIPS2018 Workshop on Meta-Learning, to take place Dec 8, papers due Oct 17! Submit at: with @ermgrant @joavanschoren Frank Hutter, Sachin Ravi.
0
3
16
@AndrewLampinen @aran_nayebi @mtoneva1 @s_y_chung And we’re done! Thank you so much to authors, speakers, panelists and participants for the turnout yesterday. We’ll be in touch about what’s next for Re-Align (!!) soon!.P.S.: This last tweet was delayed in favor of a little organizer bonding with a view, as seen below :).
0
0
15
@sucholutsky @skornblith discusses his updated understanding of representational similarity measures given experience with large language models at @Anthropic
2
1
16
The Meta-learning Workshop panel starts in <15 minutes! Ask your hard questions of the panelists (@FrankRHutter, @luisa_zintgraf, @tmh31, @LouisKirschAI, @KuanFang, Kate Rakelly) via
Meta-Learning workshop at #NeurIPS2020 is happening tomorrow! With invited speakers Frank Hutter, Luisa Zintgraf, Timothy Hospedales, Louis Kirsch, Fei-Fei Li, and Kate Rakelly. Please submit your questions for the panel here: .
0
4
15
📣 Reminder that the submission deadline is tonight, AoE!!.
The 5th edition of the Meta-Learning Workshop (#MetaLearn2021) is taking place on Workshop Monday (13th Dec.) @ #NeurIPS2021, and the CfP is now out! Please submit your up-to-8-page research papers by Sept. 17th; more details at
0
3
14
#NeurIPS2022 Outreach Day is an initiative to open up an academic venue to the local community & an opportunity for scientific communication about AI/ML to young people. They're still looking for volunteers—please sign up if you're in person @NeurIPSConf!
NeurIPS is creating a new high-school outreach day in New Orleans on Monday, Nov 28. Our vision is to empower students to engage with AI and how it impacts their lives. Discover how you can get involved:.
0
2
13
Do you know / Are you a junior researcher working on meta-learning (broadly) who could benefit from a guided introduction to the review process? Nominate them / yourself to join MetaLearn @ #NeurIPS2020 as a programme committee member!.
0
3
13
@AndrewLampinen demonstrates that careful (local/global structure-preserving) adjustments of neural network representations to fit human odd-one-out judgements improves semantic interpretability and performance on machine learning tasks, in particular in low-sample settings! 🌍
1
2
13
I sat in on last year’s AC SS held at @GatsbyUCL, and it was unlike any other SS I’ve been to: Interactive, expert intros to old & new frameworks for formalizing psychological theories. (ANNs & beyond!) I’m sure this year will be at least as good—just look at that speaker list!.
📣 Join us in Analytical Connectionism our summer school on the foundations of higher-level cognition. Join experts from neuroscience, psychology & ML for lectures, tutorials, and a hackathon! Aug 12-23, 2024 in NYC @FlatironInst @nyuniversity [ 🐭🤖.
0
0
12
@mtoneva1 uses perturbation-based interventions to demonstrate what components of linguistic data are driving the (biological) neural predictivity of language models (including large ones!): largely syntax, but also semantics (like movement), phonetics, and more ☀️
1
2
11
@s_y_chung decomposes (scalar) neural prediction error to distinguish between models that are equally predictive of neural data for different (representational) reasons. A spectral analysis concisely captures such differences via the task alignment + the intrinsic model spectrum!
1
2
9
Such neat notes! Also tagging organizer (and builder!) @FrancescoVisin and invited speakers @tmh31, @drfeifei & @KuanFang, and Kate Rakelly (.
🥳I had a great day at the #NeurIPS2020 Meta-Learning workshop 🎤 which included listening to @FrankRHutter, @luisa_zintgraf & @LouisKirschAI. Checkout many more fantastic talks & the panel! Thanks to the organizers 🤗. 🖥: .🎥:
0
0
9
We are also continuing the reviewing mentorship program, which was spearheaded last year by @janexwang: Junior & senior reviewers collaborate on a review of each assigned paper over several weeks before feedback is sent to authors. Details & sign-up at
1
1
10
@timnitGebru @le_roux_nicolas @red_abebe @mekkaokereke @isbellHFh @EricaJoy @ruha9 @shaft Thank you, @timnitGebru--I have learned a lot about allyship by listening to you speak out. Thank you for taking on that burden.
0
0
9
@judyhshen As someone who recently paid off their student loans with grad school internship money, I feel this. Lower Canadian education costs + no interest on loans during full-time grad school studies made it feasible for me. Does such a path exist for US students w/o financial security?.
2
0
9
(1) The deadline for the Re-Align Workshop at @ICLR_Conf ( is extended until Thursday, Feb. 8th, 11:59 PM AoE! 🎉 The extension is in large part due to our fantastic, numerous, diverse program committee (who make it easy to assign papers in short order).
1
2
7
We're excited to hear from Carlo Ciliberto (, @Rosemary_Ke, @Luke_Metz, @MihaelaVDS, @Eleni30fillou, Ying Wei ( on their influential work in meta-learning and related topics (data efficiency, AutoML, causal learning).
1
0
8
@sucholutsky @gpogoncheff @jacob_granley @dlipshutz argues for stochastic and dynamical similarity measures to accurately capture the function of neural responses, which are intrinsically noisy and dynamical
1
1
8
@cogphilosopher presents a free lunch for predictivity and specificity in neural comparisons: By dealing with activations in a neural network appropriately wrt their biological interpretation, we can find measures that are both predictive of, and identified by, neural data 🔥
1
3
7
@sucholutsky @skornblith @Anthropic @gpogoncheff and @jacob_granley present a fascinating natural experiment: what does early visual cortex of an unsighted human represent, and how can image-computable models help us answer this question?
1
2
7
I'm enjoying the interdisciplinary nature of this discussion on meta-learning! Please join / ask questions on the RocketChat
join us for what is an exciting Q&A panel discussion on learning to learn with our awesome speakers! Jennifer Raymond, @ermgrant, @egrefen, @janexwang, Ishita Dasgupta, Risko Miikkulainen, we are @iclr_conf !!.
0
0
7
@sucholutsky @skornblith @Anthropic @gpogoncheff @jacob_granley Lots of discussions at the morning posters!
1
0
7
The #MetaLearn2021 organizing team is @FerreiraFabioDE, myself, @FrankRHutter, @SchwarzJn_, @JoaVanschoren & @HuaxiuYaoML; get in touch via metalearn2021@googlegroups.com!.
0
0
7
@sucholutsky @gpogoncheff @jacob_granley @dlipshutz Panel on measuring representational alignment, mod @ProfData panel @jenellefeather @SarahLizHarvey @skornblith: Can we be more explicit about what the use of specific measures imply about what we're treating as important for measurement? Are stimuli and model class big confounds?
1
1
7
Biological learning dynamics via deep linear nets! Very cool work from @LiebanaSamuel et al.
How do animals learn from naïve to expert? What role does dopamine play? Find out in our new paper by @LiebanaSamuel @aeronlaffere @toschi_chiara , Louisa Schilling, Jacek Podlaski, @matfritsche, Peter Zatka-Haas, @yulonglilab, Rafal Bogacz, @SaxeLab (1/5)
0
0
6
Can confirm this is expert advice on making delicious bread 😋 🍞.
Hi! I wrote down everything I know about making sourdough in a lit review: .Theres so much out there on sourdough, this is just a little guide getting started & what matters. If you start today, you can have a starter by Friday and bread by next Sunday!!
0
0
6
@sucholutsky introduces the workshop as stemming from the Re-Align paper 🥳 and gives a shout-out to our remote presenters; you can check out their work (videos and/or posters) now at now! ⬅️
1
0
5
@ishika3singh Thanks for having me! It was great to chat with you and the other members of @bcs_IITK :).
1
0
5
An awesomely detailed thread on @artix41 and @WehenkelAntoine's contributed talk on likelihood-free inference at the meta-learning workshop.
And now @WehenkelAntoine and @artix41 presenting their work on « Recurrent machines for likelihood-free inference » #MetaLearning #NeurIPS2018
0
0
5
If you're specifically applying for a Ph.D. in machine learning (or EECS more generally), graduate students at many schools have implemented a feedback process for higher-degree apps! (Many such programs have closed, but some allow late submissions.) ⬇️⬇️.
We're running an Equal Access to Application Assistance (EAAA) program for this fall's application cycle to higher-degree programs (MS/PhD) at @Berkeley_EECS!. Any applicant to @Berkeley_EECS can submit their application packet for review by Nov. 1st at
0
1
4
@kiragoldner Not yet overwhelmed, but here are my prevention tactics in order of usefulness to me!.Keyboard shortcuts (for Gmail). Snooze till action time. Almost Inbox Zero: today-relevant emails in inbox; all others snoozed, archived or deleted. Email SEO like so:
What's it called when you search your email for something and then find the answer somewhere else and email it to yourself for next time and include all the phrase you were using to search for it? Self-SEO? Search Email Optimization?.
0
0
4
Starting in a few minutes! Zoom in to chat with @RTomMcCoy about investigating inductive biases in human language learning using meta-learning, especially if you'd like a longer discussion than during the 5min Q&A!.
Tomorrow (Aug 1) at ~10:00 AM at #CogSci2020: Come hear about using meta-learning to give linguistic inductive biases to neural networks! Track #6. We'll also have a 1-hr Zoom room for Q&A, from 10:40 AM to 11:40 AM: Paper:
0
0
4
An amazing ML for climate science opportunity at @oxcsml!.
Interested in large language models? Worried about impacts of climate change? Come join us @oxcsml @NatureRecovery @UniofOxford in pushing the frontiers in #LLMs and at the same time help #NatureRecovery and address the impacts of #ClimateChange! .
0
0
4
📩 Q's about the workshop to the Google Groups email found on and check out @sucholutsky describing the workshop in the thread below, on behalf of the organizing team: @sucholutsky @lukas_mut @khermann_ @achterbrain + 1.
🚨Call for Papers🚨 .The Re-Align Workshop is coming to #ICLR2024 . Our CfP is finally up! Come share your representational alignment work at our interdisciplinary workshop at @iclr_conf: 1/4.
1
0
4
@UpdatTheWeights @thirdreviewer I would really like to be working in the infinite data/capacity setting that the reviewer has access to---it would make a lot of these pesky problems with inductive bias go away. But then I'd be out of a job.
0
0
3
One key aspect of the workshop that we are working to enrich is virtual mentorship. In prior years, @WiMLWorkshop has hosted rotating mentorship roundtables & a more casual social; please stay tuned for a survey where you can share feedback on how to adapt to a virtual format!.
0
1
3
I highly recommend working with Tom!.
My colleagues and I are accepting applications for PhD students at Yale. If you think you would be a good fit, consider applying! Most of my research is about bridging the divide between linguistics and artificial intelligence (often connecting to CogSci & large language models).
0
0
3
@rohitrango @thirdreviewer I think this statement is a little too broad (since even some dataset biases are hard-to-detect), and some algorithms exacerbate dataset biases ( so the two are difficult to disentangle.
1
0
3
. @CS_Cornell's Student-Applicant Support Program.
Are you an underrepresented student applying to the Ph.D. program in Computer Science at Cornell? If so, please reach out to the Cornell CS Student-Applicant Support Program for advice while navigating the application process.
1
0
3
@sydgibs @BeccaRoelofs Not that my parents don't work hard (they certainly do) but their "time off" has clear boundaries, while my "time off" is an inconstant state of mind.
0
0
2
@StephaneDeny @neu_rips @automl_conf does! And maybe some of the other recent "offshoot" conferences do as well.
0
0
2
Questions to <eaaa@berkeley.edu>. EAAA is a student-led efffort, run in collaboration with @natolambert.
Excited to be launching a program at UC @Berkeley_EECS to make sure all grad applicants have access to application assistance. The student-led equal access to application assistance (EAAA) program is here: Thanks for helping us get the word out.
1
0
2
@jessebett @james_r_lucas Paging @katherine1ee, our local sourdough and scallion pancake enthusiast :).
1
0
2
@sydgibs @BeccaRoelofs Canadian, but my parents have high school diplomas and vocational training (healthcare / emergency response). I think the hardest thing to communicate has been that work/life balance is very difficult to manage when working harder does make climbing the academic ladder easier.
1
0
2
EAAA@Berkeley EECS thanks other equal-access application review programs for the originating initiative, including CS @Stanford's Student-Applicant Support Program (SASP)
We’ve started a Student-Applicant Support program for underrepresented students. If you’re considering applying to the Computer Science PhD program at Stanford, we will do our best to give you one round of feedback on your application. Apply by October 31.
1
0
2
@fhuszar @nsaphra @WiMLworkshop Added options—thank you for the feedback & for considering participating!.
0
0
1