Sign in

Atlas Wang

@atlaswang.bsky.social
2K followers 3.3K following 131 posts

www.vita-group.space 👨‍🏫 UT Austin ML Professor (on leave) www.xtxmarkets.com 🏦 XTX Markets Research Director (NYC AI Lab) Superpower is trying everything 🪅 Newest focus: training next-generation super intelligence - Preview above 👶

PostsRepliesMedia
Atlas Wang @atlaswang.bsky.social · 25/09/2026
Proud advisor face! lnkd.in/p/ef9GMXCT
lnkd.in
#thrilled #thrilled #thrilled | Atlas Wang
I’m #thrilled to see my PhD student #thrilled to share something other than “#thrilled to have K papers accepted at NeurIPS 2026!” Personally, for record I’d happily sign off on a PhD dissertation on...
010
Reposted by Atlas Wang
GIGAZINE @gigazine.net · 28/05/2026
「AIも使い続けていると人間のように老化する」という指摘、セッションを重ねて記憶が蓄積されることで性能悪化 gigazine.net/news/20260528-agingben…
gigazine.net
「AIも使い続けていると人間のように老化する」という指摘、セッションを重ねて記憶が蓄積されることで性能悪化
AIエージェントはどれだけ使用しても人間のように疲れることがなく、加齢などによる認知機能の低下もないと考えられています。ところが、テキサス大学オースティン校の研究チームが「AIエージェントも老化する」と指摘し、AIエージェントの老化を測定するベンチマークの「AgingBench」を発表しました。
02718
Reposted by Atlas Wang
utcompsci.bsky.social @utcompsci.bsky.social · 19/05/2026
Congratulations to UT Computer Science Ph.D. student Yeonju Ro on being named a 2026 MLCommons ML and Systems Rising Star! Co-advised by Aditya Akella and @atlaswang.bsky.social, Yeonju is also a contributor to UT’s @ldosexpedition.bsky.social and Infra AI Center.
073
Reposted by Atlas Wang
WIRED @wired.com · 22/10/2025
A new study shows that feeding large language models low-quality, high-engagement content from social media lowers their cognitive abilities. www.wired.com/story/ai-mod...
wired.com
AI Models Get Brain Rot, Too
A new study shows that feeding large language models low-quality, high-engagement content from social media lowers their cognitive abilities.
2627484
Atlas Wang @atlaswang.bsky.social · 21/10/2025
LLMs can get “brain rot”! www.linkedin.com/posts/atlas-...
linkedin.com
Pleasantly surprised to see a side project, co-led by the now-professor Junyuan Hong and sparked by a totally random idea myself threw out back in January, gaining lots of (organic) traction on… | Atl...
Pleasantly surprised to see a side project, co-led by the now-professor Junyuan Hong and sparked by a totally random idea myself threw out back in January, gaining lots of (organic) traction on Twitte...
041
Reposted by Atlas Wang
Conference on Parsimony and Learning (CPAL) @cpalconf.bsky.social · 23/09/2025
Calling all parsimony and learning researchers 🚨🚨 The 3rd annual CPAL will be held in Tübingen Germany March 23–26, 2026! Check out this year's website for all the details cpal.cc
031
Atlas Wang @atlaswang.bsky.social · 18/09/2025
Got a comment I secretly like: “This professor is constructively rude”
010
Atlas Wang @atlaswang.bsky.social · 02/09/2025
www.linkedin.com/posts/atlas-...
linkedin.com
Job Description | XTX Markets | Atlas Wang
🌐 It has been an incredible first year for the XTY AI Lab of XTX Markets in New York City, where we set out to build a world-class hub for AI research in quantitative finance: the results have exceede...
020
Atlas Wang @atlaswang.bsky.social · 31/08/2025
Paying for a dinner I wasn’t even at 🤣 — momentum’s strong, keep it up team! www.vita-group.space/team
140
Atlas Wang @atlaswang.bsky.social · 27/08/2025
I really wanna see / do research on using AI to prevent massive shooting …
010
Atlas Wang @atlaswang.bsky.social · 08/08/2025
Perhaps I’m not alone in feeling GPT5 (even pro) is no better than o3-pro… coding and knowledge?
110
Atlas Wang @atlaswang.bsky.social · 08/08/2025
www.linkedin.com/posts/atlas-... Cannot resist enjoying some proud advisor moment
linkedin.com
Eight years ago, I walked into my faculty office for the first time, not knowing what the journey ahead would hold. | Atlas Wang | 10 comments
Eight years ago, I walked into my faculty office for the first time, not knowing what the journey ahead would hold. This past year, while I’ve been gradually “fading into the background,” the group ...
041
Atlas Wang @atlaswang.bsky.social · 08/07/2025
Two fresh tutorials on low-dimensional DL: Sparsity ’23 🔗 vita-group.github.io/SparsityTuto... Low-Rank ’25 🔗 vita-group.github.io/LowrankTutor... Curious about how sparsity and low-rank ideas power modern deep learning? I hope these slide decks & notes are a useful starting point! 🚀
vita-group.github.io
Sparse Neural Networks: From Practice to Theory | Atlas Wang
0112
Atlas Wang @atlaswang.bsky.social · 03/07/2025
Of all my students’ career achievements, this one makes me proudest: because really, who needs yet another AI professor or scientist alumni when you can boast a professional cat-breeder student? www.bestbritishcats.com To my Bay Area friends: BUY one! Maybe try using my name to get a discount :)
bestbritishcats.com
Snow Moon British | 🌴 CA, 🇺🇸 | Affectionate British Kittens for You
Snow Moon is a trusted British cattery located in San Jose, CA, specializing in healthy, champion-quality British Shorthair and Longhair kittens. Ethical breeder. Breeding cats are genetically tested,...
010
Atlas Wang @atlaswang.bsky.social · 30/06/2025
@talkachman.bsky.social caught it faster before I did... but here we go! 🚀 New pre-print dropped lnkd.in/gmQ9hBdf Its first draft won the #DARPA Disruptive Idea award at #NeuS2025. Now my student Peihao Wang has sharpened the theory even more. I’m so excited what it means for AI reasoning 🧩🤖 (1/n)
lnkd.in
LinkedIn
This link will take you to a page that’s not on LinkedIn
120
Reposted by Atlas Wang
talkachman.bsky.social @talkachman.bsky.social · 30/06/2025
@atlaswang.bsky.social knocking out of the park again! half a page in and this is a phenomenal read. So much to wrap my head around www.arxiv.org/abs/2506.21797
arxiv.org
Why Neural Network Can Discover Symbolic Structures with Gradient-based Training: An Algebraic and Geometric Foundation for Neurosymbolic Reasoning
We develop a theoretical framework that explains how discrete symbolic structures can emerge naturally from continuous neural network training dynamics. By lifting neural parameters to a measure space...
211
Atlas Wang @atlaswang.bsky.social · 26/06/2025
Feels great relief to report: #Thrilledtoshare (yes, that LinkedIn tone I hate...) we have 0 papers accepted to @iccv.bsky.social ! Surprisingly liberating to step off the number game: I meant it. I’m genuinely proud of the science my students are building, no matter how ICCV reviewers think 🤣
060
Reposted by Atlas Wang
JAMA @jama.com · 23/06/2025
An #AI system that automatically interprets echocardiograms maintained high accuracy across geography and time from complete and limited studies. ja.ma/4jYWBgI
Figure 4. Task-Specific ViewRelevance
194
Reposted by Atlas Wang
Evangelos K. Oikonomou @ekoikonomou.bsky.social · 25/06/2025
Check out #PanEcho, now in @jama.com ✅ open weights ✅ multiview multitask reporting ✅ international validation ✅ works with POCUS Amazing effort @cardslab.bsky.social @giholste.bsky.social @rohankhera.bsky.social + @atlaswang.bsky.social, Márton Tokodi & Attila Kovács #CardioSky #EchoSky
184
Atlas Wang @atlaswang.bsky.social · 15/06/2025
A reflection sparked by Saturday night’s thought-provoking dinner conversation 🥘 www.linkedin.com/pulse/beyond...
linkedin.com
Beyond Bigger Models: AI Needs Bigger Problems
(Disclaimer: AI helped revise this article, too: -) “If AI is truly transformative, why are its dividends still so concentrated?” Over the past two years, I have watched two headlines run in parallel:...
060
Atlas Wang @atlaswang.bsky.social · 11/06/2025
Making a new website of my research group, and did a visualization of all our papers from 2018 to present: clustered into 10 topics. One can clearly see how this group evolves its own tastes! … and deeper in my heart: long live optimization!! ❤️
030
Atlas Wang @atlaswang.bsky.social · 25/05/2025
Quoting one slides from @yann-lecun.bsky.social talk… arxiv is filled by papers that treat symptoms (or not even!) without ever diagnosing the disease
043
Atlas Wang @atlaswang.bsky.social · 25/05/2025
Unwelcome idea: I’m extremely tired of seeing one more LLM paper and believe academics shall pivot away from this area. So boring arxiv feeds every day…
130
Reposted by Atlas Wang
Matt Peeples @pattmeeples.bsky.social · 02/05/2025
This is going to kneecap science in this country for years. www.nature.com/articles/d41...
nature.com
Exclusive: NSF stops awarding new grants and funding existing ones
US science funder also plans to screen grant applications for compliance with ‘agency priorities’.
153
Reposted by Atlas Wang
Yisong Yue @yisongyue.bsky.social · 19/04/2025
One of my PhD students got their visa revoked. I know of other cases amongst my AI colleagues. This is not what investing in US leadership in AI looks like. www.aljazeera.com/news/2025/4/...
aljazeera.com
US revokes nearly 1,500 student visas: Who are the targets?
Hundreds of students have had their visas cancelled and find themselves in limbo.
25923
Reposted by Atlas Wang
Atlas Wang @atlaswang.bsky.social · 18/03/2025
🚀 Thrilled to announce SPIN-Bench!🚀 We all love seeing how smart LLMs can be-solving complex math, crafting beautiful text, and coding effortlessly. But how well do they handle real-world strategic complexity, cooperation, & social negotiation? Can they play well when things get tricky? Not quite!
181
Atlas Wang @atlaswang.bsky.social · 27/03/2025
Just had a meal that gifted me two rare treasures: 1️⃣ Meeting someone infinitely wiser than me. 2️⃣ They weren’t cold or mean—just gently showed me where I could grow. "To learn truth at dawn, I’d die content by dusk." ✨ Humility tastes better with kindness. #Gratitude #LifeLessons
030
Reposted by Atlas Wang
Atlas Wang @atlaswang.bsky.social · 18/03/2025
Good news? SPIN-Bench pinpoints exactly where these models fall short—and illuminates exciting research directions to smarter, socially savvy AI For some extra fun (and detailed, interactive trajectory visualizations!), visit project page: 👉 spinbench.github.io full paper: arxiv.org/pdf/2503.12349
spinbench.github.io
SPIN-Bench: How Well Do Large Language Models Reason Strategically and Socially?
How Well Do Large Language Models Reason Strategically and Socially?
021
Reposted by Atlas Wang
Atlas Wang @atlaswang.bsky.social · 18/03/2025
🌍 Diplomacy – The ultimate test! Models had to negotiate, forge alliances, and occasionally backstab. Result? Even the best LLMs floundered, struggling to juggle complex social interactions and strategic depth.
121
Reposted by Atlas Wang
Atlas Wang @atlaswang.bsky.social · 18/03/2025
🎴 Cooperative Games (Hanabi): Coordination among teammates dramatically challenges models, causing their performance to dip sharply as complexity ramps up. Turns out, keeping track of your teammates’ intentions isn’t an easy task—even for GPTs!
111
Reposted by Atlas Wang
Atlas Wang @atlaswang.bsky.social · 18/03/2025
Results fascinatingly reveal: 🔹 Classic Planning: LLMs ace simpler puzzles but struggle badly as complexity grows—losing track at longer-term decisions 🔎 Competitive Games: Top chess engines swept every LLM clean. Even simple tactical awareness quickly fades when facing deeper strategic branches
121
Atlas Wang @atlaswang.bsky.social · 18/03/2025
This accidental photo feels distinctly #american —industry and technology advancing, sometimes hesitantly, beneath the weight of religion and firearms
010
Atlas Wang @atlaswang.bsky.social · 18/03/2025
🚀 Thrilled to announce SPIN-Bench!🚀 We all love seeing how smart LLMs can be-solving complex math, crafting beautiful text, and coding effortlessly. But how well do they handle real-world strategic complexity, cooperation, & social negotiation? Can they play well when things get tricky? Not quite!
181
Reposted by Atlas Wang
Katie Mack @astrokatie.com · 28/01/2025
I think some people hear “grants” and think that without them, scientists and government workers just have less stuff to play with at work. But grants fund salaries for students, academics, researchers, and people who work in all areas of public service. “Pausing” grants means people don’t eat.
washingtonpost.com
White House pauses all federal grants, sparking confusion
The Trump administration has put a hold on all federal financial grants and loans, affecting tens of billions of dollars in payments.
15754324114330
Atlas Wang @atlaswang.bsky.social · 23/01/2025
Since taking absence from my university role, students thrive even better: consistent conference productivity (NeurIPS ICLR etc basically on autopilot); two new PhD fellowships (NVIDIA, IBM); multiple paper awards; and most recently, several successful job placements including a new professor (HKU)
170
Atlas Wang @atlaswang.bsky.social · 20/01/2025
I love this adorable lady’s narrative and hope it seen by more people 【Asians being seen 第二则-哔哩哔哩】 b23.tv/GZ4h0HO
b23.tv
Asians being seen 第二则_哔哩哔哩_bilibili
Asians being seen 第二则, 视频播放量 1655、弹幕量 2、点赞数 293、投硬币枚数 72、收藏人数 19、转发人数 1, 视频作者 ItsNOTViola, 作者简介 ig- itsnotviola |viola98finn@outlook.com|全平台同名|在d.y上(有时)直播,相关视频:Am I a hater? 我的第一则声明,洋抖实力依旧,谷爱凌的精力与她的pr...
000
Atlas Wang @atlaswang.bsky.social · 13/01/2025
current price to buy or invest a #GenAI startup: is it already too inflated?
040
Atlas Wang @atlaswang.bsky.social · 08/01/2025
Me: I’m feeling like my life on a busy shaky ferry😭 My junior student: Maybe a busy shaky ferry in new waters is better than a calm ferry always in the same waters #IAmEducated
080
Atlas Wang @atlaswang.bsky.social · 06/01/2025
medium.com/@brawlingoce...
medium.com
No, A woman did not quit her PhD to do OnlyFans and you shouldn’t either
Zara Dar claims she quit her PhD in order to pursue OnlyFans fulltime
040
Reposted by Atlas Wang
Atlas Wang @atlaswang.bsky.social · 24/12/2024
The not-so-great-yet human intelligence wishes y’all Happy Holidays! 🎁🎄
0161
Reposted by Atlas Wang
talkachman.bsky.social @talkachman.bsky.social · 29/11/2024
Hard at work on the AAMAS rebuttals. Yes reviewer 2 we.are looking at you
261
Atlas Wang @atlaswang.bsky.social · 24/12/2024
The not-so-great-yet human intelligence wishes y’all Happy Holidays! 🎁🎄
0161
Atlas Wang @atlaswang.bsky.social · 22/12/2024
With o3 rocketing to (kinda) super-intelligence overnight, I’m prepping my fallback gig: selling “hand-crafted” research to anyone nostalgic for the pre-AI era. Turns out boutique history may be the only job left! #FutureProof #AIRevolution
070
Reposted by Atlas Wang
Melanie Mitchell @melaniemitchell.bsky.social · 21/12/2024
Excellent post about the recent OpenAI o3 results on ARC (& other benchmarks). I don't know how @natolambert.bsky.social manages to write these so quickly! I highly recommend his newsletter. www.interconnects.ai/p/openais-o3... I am (more slowly) writing my own take on all this, coming soon.
interconnects.ai
o3: The grand finale of AI in 2024
A step change as influential as the release of GPT-4. Reasoning language models are the current big thing.
215231
Atlas Wang @atlaswang.bsky.social · 19/12/2024
@ruisicai.bsky.social was awarded NVIDIA graduate fellowship blogs.nvidia.com/blog/graduat... It marks the 8th PhD fellowship brought back to VITA group in the last 4 years🎇🎆 alongside NSF GRFP, Apple, Amazon, Adobe, IBM, Qualcomm, and Snap When students are too compelling, PI feels useless :(
blogs.nvidia.com
NVIDIA Awards up to $60,000 Research Fellowships to PhD Students
The NVIDIA Graduate Fellowship Program announced awards for Ph.D. students involved in research that spans all areas of computing innovation.
0130
Reposted by Atlas Wang
Atlas Wang @atlaswang.bsky.social · 10/12/2024
(5/n, n =5) More cool APOLLO videos for seeing with your own eyes 👀 AdamW with micro-batch size of 4 at 79-80GB memory cost youtu.be/XuNTX7VDViQ APOLLO with micro-batch size of 16 (four times larger!) at 69-70GB memory cost youtu.be/s2WxdBGh99k 🤩 Codes will be fully available soon enough!
youtu.be
Pre-train a LLaMA 7B on 8×A100-80GB with Memory-Efficient APOLLO
YouTube video by Hanqing Zhu
141
Reposted by Atlas Wang
Atlas Wang @atlaswang.bsky.social · 10/12/2024
(1/n) My favorite "optimizer" work of 2024: 📢 Introducing APOLLO! 🚀: SGD-like memory cost, yet AdamW-level performance (or better!). ❓ How much memory do we need for optimization states in LLM training ? 🧐 Almost zero. 📜 Paper: arxiv.org/abs/2412.05270 🔗 GitHub: github.com/zhuhanqing/A...
2154
Atlas Wang @atlaswang.bsky.social · 10/12/2024
Everyone’s saying #GenAI tools ‘unleash human creativity’— the new politically correct mantra! But are they empowering us, or just tricking us into creating better training data for themselves? 🤔 Who’s really learning here? #HumanVsMachine
010
Atlas Wang @atlaswang.bsky.social · 10/12/2024
fortune.com/2024/12/09/n... It’s exciting to see another pitch on neurosymbolic AI by @fortune.com , after one nice by @reuters.com last month I don’t believe we are already in the monotonic deceasing “convex optimization”-like path of AI. It’s always more fun to back & forth like good non-convex!
fortune.com
Why neurosymbolic AI could be the next big thing
Neurosymbolic AI could be a best-of-both-worlds marriage between deep learning and “good old-fashioned AI.”
041
Atlas Wang @atlaswang.bsky.social · 10/12/2024
(1/n) My favorite "optimizer" work of 2024: 📢 Introducing APOLLO! 🚀: SGD-like memory cost, yet AdamW-level performance (or better!). ❓ How much memory do we need for optimization states in LLM training ? 🧐 Almost zero. 📜 Paper: arxiv.org/abs/2412.05270 🔗 GitHub: github.com/zhuhanqing/A...
2154