Atul Prakash

63 posts

Atul Prakash

Atul Prakash

@atulprakash

Katılım Ocak 2009
103 Takip Edilen120 Takipçiler
Atul Prakash retweetledi
MichiganAI
MichiganAI@michigan_AI·
Big congratulations to @ziqiao_ma for winning the Towner Prize for Outstanding PhD Research.🎉 It's a major recognition of creativity, impact, and outstanding contributions to #AI research. A well-deserved recognition!👏 cse.engin.umich.edu/stories/martin…
English
0
4
18
5.7K
Atul Prakash retweetledi
Michigan Human-Centered Computing Lab
Congrats to Farnaz Jahanbakhsh & Rayhan Rashed on winning the 🏆 UM Social Impact Engineering Award!🏆 Their #CHI26 work proposes a new paradigm for online moderation: transform content to preserve value while reducing harm — personalized to each user. rayhan.io/diymod/
Michigan Human-Centered Computing Lab tweet media
English
2
5
28
1.4K
Atul Prakash retweetledi
MichiganAI
MichiganAI@michigan_AI·
🎙️Thrilled that Episode 1 of the Deep Blue | The Michigan AI Podcast is here! Vaibhav Balloli @v_balloli interviews Prof. Alex Rodriguez @_arodriguezca on using #AI for Uncertainty: Predicting epidemics and disasters:⬇️ youtube.com/watch?v=U2grAX…
YouTube video
YouTube
English
0
2
3
589
MichiganAI
MichiganAI@michigan_AI·
Don't miss this #SIAMDA26 paper! Shabarish Chenakkod, Michał Dereziński, & Xiaoyu Dong give a proof of the conjecture of Nelson & Nguyen [FOCS 2013] on the optimal dimension and sparsity of oblivious subspace embeddings, up to sub-polylogarithmic factors: epubs.siam.org/doi/10.1137/1.…
MichiganAI tweet media
English
1
1
1
189
Atul Prakash
Atul Prakash@atulprakash·
@michigan_AI Dimensionality reduction is so important for efficient machine learning! Congrats on proving the conjecture.
English
1
0
2
16
Atul Prakash
Atul Prakash@atulprakash·
@Delta I fly delta a lot and it is my favorite airline. But who thought of this new design of pockets with holes on the sides? This is on a A330 from Detroit to Amsterdam.
Atul Prakash tweet media
English
0
0
1
40
Atul Prakash
Atul Prakash@atulprakash·
See our new work, FOCAL, at ICML 2025 for adding robust perception at test-time to vision foundation models such as CLIP under lighting shifts, rotations, etc. without retraining the foundation model. Congrats @ryantfeng @utksinghal and Stella Yu.
Ryan Feng@ryantfeng

Q: How do we scale robustness/invariance to foundation models like CLIP? A: Test-time search! 🔍 Our new work FoCal finds canonical views to boost robustness to complex transforms (e.g. viewpoint): sutkarsh.github.io/projects/focal 📍 ICML Poster: Tue 11–1:30, E. Hall A-B (E-2203) 🧵 1/5

English
0
2
6
992
Atul Prakash
Atul Prakash@atulprakash·
Congratulations @Maggiemakar on winning the 2024 Google Research Scholar Award for work titled "Reliability through causal alignment: causally motivated models for pain management." Causally-aligning ML models to make them trustworthy is critical in healthcare settings.
English
0
2
18
1.9K
Atul Prakash
Atul Prakash@atulprakash·
@EarlenceF Congratulations Earlence on the Google Research award. Getting privileges correct with OAuth-style authentication is of huge importance.
English
0
0
1
59
earlence
earlence@EarlenceF·
Excited for this! Supports our work on improving OAuth-like authorization protocols to have better least privilege properties. This is fundamental infrastructure that powers stuff ranging from IoT to your latest ChatGPT "agentic" gizmos.
UCSD CSE@ucsd_cse

Bravo to @ucsandiego CSE assistant professors @EarlenceF and @amyousterhout for being among 70 recipients for @Google's Research Scholar Program! 👏 Read more: tinyurl.com/587t2ppa #UCSD #UCSanDiego #UCSDCSE #Google #Research #ComputerScience

English
11
1
46
3.6K