XOutlier
Download .zipAdd to Chrome

Sara Hooker

@sarahookr65k followers

Building intelligence that evolves @adaption_ai. Built @Cohere_Labs, @GoogleBrain, @GoogleDeepmind. ML Efficiency, Multimodal\lingual.

Open on X
Top outlier
976×
Best organic post
Peak views
8.9M
Single post
Bookmarks
15k
Across 93 organic posts
Last scan
Oct 16
Community cache
Posts 93 posts
#DatePost↓ OutlierViewsLikesRepostsBookmarks
1Oct 30, '25Adaptable Intelligence. Multiple possible paths to an objective.976×8.9M15k1.1k3.7k
2Mar 2, '25The first step of any meaningful pursuit is to severely underestimate its difficulty.900×8.2M16k2.3k2.3k
3Feb 19, '23If you have multiple papers before you even began a PhD, it likely means you had access that others didn't. I wish more PhD programs would take a step back and stop this absurd pra93.8×855k2.9k259165
4Dec 2, '23I emailed an author about a paper where the ENTIRE contribution is about a dataset they curated + constructed. The question: has the dataset been released? No, we decided not to. W39.9×364k1.6k57113
5Apr 30, '25It is critical for scientific integrity that we trust our measure of progress. The @lmarena_ai has become the go-to evaluation for AI progress. Our release today demonstrates the d31.2×284k719135281
6Oct 25, '25Today I tried horse meat. 400% better than I imagined.28×255k440431
7May 28, '24lol someone wants to train a multi-modal model.25.8×235k1.8k7591
8Oct 7, '25I'm starting a new project. Working on what I consider to be the most important problem: building thinking machines that adapt and continuously learn. We have incredibly talent den24.7×225k2.5k1901.2k
9Feb 15, '21Yesterday, I ended up in a debate where the position was "algorithmic bias is a data problem". I thought this had already been well refuted within our research community but clearl19.7×02.6k668783
10Mar 5, '23We need more nuanced discussions around the risk of open sourcing models. Open source brings valuable access, but it is absurd to ignore the fact that it lowers the barriers to ent19.7×180k4376384
11Feb 14, '23What is your favorite matplotlib configuration setting for beautiful scientific charts? Links welcome to open source or examples of charts you love.19.1×174k75987812
12May 11, '25Following release of our recent work, we have spent considerable time engaging with @lmarena_ai over last week. The organizers had concerns about the correctness of our work on the16.5×150k54089199
13Jul 10, '24Does more compute equate with greater risk? What is our track record at predicting what risks emerge with scale? I don't have much time anymore to write, but this felt important en15.8×144k32277187
14Aug 11, '25It has been an incredible honor to spend the past few years leading @Cohere_Labs @cohere . This has been the adventure of a lifetime. However, after much deliberation, I made a tou15.2×138k1.3k20119
15Nov 12, '23lol what happened to the flurry of research on scaling laws. have we conceded by now this was a "squint hard enough to see the pattern" vibe of research13.8×126k2331782
16Dec 13, '24What has been your favorite poster at NeurIPS so far. No posters by your lab or immediate colleagues allowed lol. Let’s keep this fun. Paper links welcome.13.5×123k35728248
17Oct 14, '25Adaptive but not intelligent. Drone footage from Lior Patel.12.9×118k62332161
18Jul 22, '24Is bigger always better? 🐘 The idea that scaling more than any other ingredient has driven progress has become formalized as the “bitter lesson” Is Sutton right? 📜12.1×110k42981375
19Feb 13, '24Today, I am very proud share what we have been working on for the last 14 months. ✨ Introducing Aya -- a new state-of-art for massively multilingual models. 🔥🎉10.8×98k989157190
20Aug 8, '23We are hiring for both research scholars and research scientist roles. 🎉🔥 Come join us, and work on fundamental questions in machine learning at scale.9.4×86k36764170
21Apr 26, '24Two years ago today I joined @cohere to bring back as @CohereForAI💙 Thank you to everyone who has supported our work in the two years since. I am a bit in awe of everything that h8.4×76k3571824
22Apr 21, '19Driving through Utah near Zion... ok, you have my attention.8.3×01.8k23421
23Jan 18, '25Ok this is absurd. Dude — we are humans. Live your best life instead of limiting yourself to absurd stereotyping of what is feminine vs masculine.8.0×73k5602153
24Mar 22, '24We are hiring a machine learning engineer role to drive making our research + weight releases as accessible as possible to the wider community. 🔥 If you care about model efficienc7.9×72k30770143
25Jun 29, '24I don’t know who changed the @X algorithm — but the random rubbish in my timeline is driving me crazy. No — I don’t want constant posts about politics and memes all day from people7.9×72k474276
26Jan 5, '25So collusion rings are still a thing at top conferences. @openreviewnet has all the data across all top tier ml conferences — what are they currently do to solve this. What top tie7.6×69k2841957
27Aug 16, '25Number of recruiter inbounds in 5 days — 47😂 Just to make everyone’s life easier — I’m not looking for a new job right now. If you do still want to persevere — send me your favori7.3×67k322422
28Aug 8, '20There is a misconception that famous researchers in ML are innately brilliant. Most well known researchers I meet are not. They are smart. But the dominant traits are perseverance 7.3×01.5k16481
29Dec 21, '23Before we started @CohereForAI I asked well known researchers for advice — the biggest skepticism I met was it wouldn’t be possible to do top tier research while creating entry poi7.3×66k3293729
30Feb 23, '24Were you curious to find out @GoogleDeepMind Gemma is using REINFORCE instead of PPO? We have been working on this paper for months --that takes apart PPO and motivates REINFORCE S7.2×66k45369290
31May 3, '25We recently released a paper where I took on a more visible role than typical. This was a deliberate choice to protect more junior leads given we anticipated more scrutiny than typ6.7×61k3932638
32May 1, '23Ive arrived — 3 am and just ordered Indian chicken masala room service. I’m so glad I’m back in East/central Africa where excellent Indian food abounds.6.7×61k339510
33Jun 20, '23Exciting! We are hiring for a research scientist role. 🎉 One of the areas of interest is security/risk at scale: - rigorously evaluating model risk at scale - efficient red teamin6.1×56k2425173
34Nov 14, '24I am not sure why people are surprised to hear there is a limit to scaling. While our current architectures represent a huge step forward in performance, what is often missed is ar5.8×53k3394097
35Mar 29, '25Do you remember when "prompt engineering" was introduced as the future lol. We put all the burden on users to probe high dimensional spaces like tea leave reading. And then every t5.7×52k4241972
36Aug 8, '24Next week we open up applications for the @CohereForAI scholar program. 👀 Join us as we work on breakthroughs on the frontier of ML. More details to follow in a few days. 🔥🌎🌍🌏5.5×50k24334117
37Sep 5, '24A special moment for all the work we are doing -- thank you @TIME. ✨5.5×50k5584429
38Jan 20, '25Excited to share we are hiring for a new type of scholar role. Algorithm interface scholars will re-imagine how humans interact with algorithms. These scholars will lead the fronti5.4×49k2464482
39Jan 31, '23Happy 6 year wedding anniversary to us! 🎉5.2×48k43853
40Dec 30, '24Feel the need to point out again — even slightly more sophisticated sampling can overcome mode collapse in synthetic data.5.0×45k38655179
41Oct 4, '24One of the biggest open questions is what is the limit of synthetic data. Does training of synthetic data lead to mode collapse? Or is there a path forward that could outperform cu4.9×45k33243215
42Sep 2, '22Efficient Methods for Natural Language Processing: A Survey A survey that started in wonderful @dagstuhl and now includes the input of many institutions and authors. A great starti4.8×0614130266
43Oct 7, '25I took a leap 3 weeks ago and can't believe that today the new adventure begins. I feel the same energy I felt when we were just beginning @Cohere_Labs -- when it felt very clear t4.6×42k415930
44Apr 22, '22Today is my last day at Google Brain. My biggest takeaway is that research is an inherently human endeavor. It has been a special 5 years. I have learned a lot from all of you, and4.6×01.1k2456
45Jun 14, '22I'm excited to finally share what I have been working on. Today we are officially launching Cohere For AI @forai_ml a non-profit research lab that aims to reimagine how, where, and4.5×078811085
46Dec 29, '24Such a blast from the past. My mom sent this to me today. 2017 vs (almost) 2025, so much has changed — Google Brain was a very special era.4.4×40k263950
47Aug 10, '23Final pass of NeurIPS rebuttals done! Time for a day of scuba diving.4.1×37k29933
48Nov 21, '19What does a pruned deep neural network "forget"? Very excited to share our recent work w Aaron Courville, Yann Dauphin and @DreFrome4.0×0583153127
49Jun 3, '24Our SF office is beautiful.3.9×36k388819
50Apr 12, '21Moving beyond "algorithmic bias is a data problem" An extended perspective piece in @Patterns_CP How do our model design choices contribute to algorithmic bias?3.9×0538127150
51Nov 2, '25Today I learnt about the lava lamp wall in computer science history. It is difficult for computers to produce truly random numbers. Lava lamp "flow" is very unpredictable, so cloud3.8×35k3411867
52Feb 13, '23Makes my entire day. When you write a recommendation letter and the person gets into the program they applied to. 🔥🔥3.7×34k25235
53Feb 10, '24While millions end up using open source datasets, they only exist because of the effort of a small group of humans. It is perhaps the most undervalued contribution in open science.3.4×31k3514426
55May 21, '25A lot of people are asking me what I think about @lmarena_ai going private with 100M funding. I wish them well, the goal was never to take down the arena. However, we stand by our 3.4×31k2432736
56Nov 18, '18ICLR 2020 will be in Addis! Exciting news not only to reduce the visa issues that have plagued recent conferences, but also because I think many of my colleagues will be inspired b3.4×055515515
57Oct 8, '25Thank you to everyone who has reached out since yesterday. Trying to keep up. 🔥 We put together a form if you want to support, learn about future releases. We will also be hosting3.4×31k2771042
58Aug 16, '25This was a very welcome and high value email 🔥3.3×30k264127
59Oct 10, '25We are hiring an efficiency research engineer. I'll share a little bit about the role, because our hiring says a lot about what we stand for, and how we want to build technology. S3.2×29k31521127
60Dec 16, '24Everybody has talent, but learning a craft takes hard work. That is why it is hard to predict what researchers become great, often it is the researchers who stick with an idea the 3.1×29k2682561
62Oct 30, '25Join us. You will work on the most important problem, with an incredibly talent dense founding team. Time is limited, spend it working on something that changes what is possible.3.0×27k30927127
63Nov 17, '25octopus dreaming. rapid camouflage as the dreams simulate different environments.3.0×27k2872649
64Jun 30, '25One of the first technical reports in a while that has considerable detail. In particular -- preference training tends to be much more painful for multimodal, largely because of la2.8×26k33030292
65Oct 13, '20It is not just the data. Popular compression techniques amplify bias in deep neural networks. Work co-led with @nunuska, w @_whatcode, Samy Bengio and @cephaloponderer "Characteriz2.7×03899682
66Nov 6, '25This video speaks to my soul. If all my tabs were papers my desk would look like this.2.7×25k3891871
67Mar 10, '20Me and my brother @SeanHooker have a fun new hobby podcast. It is called @underrated_ml . We choose 3 machine learning papers/ideas each episode that haven't gotten the attention a2.7×05135357
68Jun 24, '23EMNLP submissions done. It’s Friday and time to celebrate my birthday!2.6×24k27851
69Jul 2, '25Now officially a citizen of both Ireland and the US 🇮🇪 🇺🇸 Had my naturalization ceremony last week after a long journey as an immigrant in the US, very special to finally be do2.6×24k481318
70May 23, '25I came to the US as an international student and it changed my entire life — opened so many doors. Harvard ban will impact many US schools as international students will be uncerta2.5×23k265158
71Jun 23, '21How do hardware, software and algorithm contribute to non-determinism in deep neural networks? What is the downstream impact and the cost of ensuring determinism? Very excited to s2.4×03517982
72Aug 14, '24This is such a special moment -- Aya was as much about changing how research was done as it was about a scientific breakthrough. ✨ Thank you to @aclmeeting for recognizing our work2.3×21k2633116
73Jun 9, '25Second night — celebrating at the hard to reserve Tamil Prince. Now my love of Indian food is known — I am getting a tour of all the best spots. Summer in London is fabulous.2.3×21k257429
74Oct 24, '22I'm hiring for a research scientist position studying efficiency at @forai_ml -- how can we train effective representations w fewer resources. Making ML more accessible by minimizi2.3×03886168
75Oct 22, '25Taking a founding team photo today. We have to do at least one version like this for the vibes.2.3×21k271512
76Jan 18, '25Honestly the current trend from well known silicon valley leaders to ascribe masculine and feminine energy to positive and negative attributes is the most basic vibe I’ve seen sinc2.2×20k287138
77Dec 2, '21Someone who regularly puts the spotlight on others is @hardmaru. I really appreciate all the small ways (often behind the scenes) he supports other researchers + invests the time t2.2×0557198
78Oct 20, '25Happy diwali to all who celebrate 🪔✨2.1×19k51394
80Mar 11, '20this is amazing.2.0×04236311
81Dec 9, '19Presenting "A Benchmark for Interpretability Methods in Deep Neural Networks" at poster #159 tomorrow (Dec 10th). Poster > Slides > Code > Paper > See you there! :)1.9×02657660
82May 7, '22I have been thinking a lot about how we can change how, where and by whom research is done. how do we provide more entry points into ML research than a single path from undergrad-&1.8×03352060
83Jul 23, '21How do you distinguish between sources of uncertainty? This is important because the downstream remedies for atypical and noisy examples are very different. Two of our workshop pap1.8×02623588
85Jun 23, '24Feeling very special and loved on my birthday. Thank you friends close and far 🎉💙1.7×15k28923
86Nov 6, '17Our work shows the explanations of a model prediction can be mislead by simple input transformations to show a cat.1.6×0268961
87Jun 22, '18It’s official! I’ll be joining @Moustapha_6C at the new @GoogleAI research center in Accra, Ghana this year. Excited to be part of this new research initiative and see the ML ecosy1.6×0349291
88Nov 23, '21Wow! The Hardware Lottery is the December cover of @CACMmag. This is really special, grateful this work has resonated with so many. Thanks to all, including @vardi who nudged me to1.6×02894520
89Jun 16, '22It has arrived! @chipro — all the more special knowing how hard you worked on this. 🎉🔥 So proud of you!!1.4×02832041
90Sep 24, '22Grumpy take -- this is an absurd way to begin a scientific discussion. These numbers are not grounded in anything. It is akin to saying: "I believe aliens are arriving in 10 days b1.3×02521521
92May 14, '21So exciting to see more people work on the relationship between compression + generalization. When I started working on the "what do compressed deep neural networks forget?" paper 1.3×02402040
94Mar 21, '25The end of an amazing week in Patagonia. It has been a long time since I’ve been completely offline — has been very worth it.1.2×11k25815
95May 18, '20This is absolutely awful. I was on OPT — the US will inevitably lose out on smart students from all over the world. It’s just absurd.1.1×0234334
96Jul 12, '22Is this what happens when conferences are held in Seattle? Everything looks like a farmers market instead of a cupcake table at coffee breaks.1.0×024079
97Jun 14, '22Also, so excited I can finally share that I have joined @CohereAI!! So grateful to have joined this wonderful group of humans. ❤️❤️ @aidangomezzz @nickfrosst @1vnzh @egrefen @pblun1.0×0245134
98Dec 31, '21Happy new years everyone! 🎉 The highlight of my year happened yesterday. I checked the status of my US green card application and... Fingers crossed early next year I will finally1.0×025413
99Dec 28, '20I’m sorry I wrote you such a long letter; I didn’t have time to write a short one. — Blaise Pascal.0.9×023088

Scan another creator

@
@karpathy@levelsio@paulg@naval@marclou@illyism

Run this on any account.

XOutlier builds this table for every public profile you open on x.com, with live data and every post, not just the top 50.

Add to Chrome Download .zip