Jingfeng Wu

123 posts

Jingfeng Wu banner
Jingfeng Wu

Jingfeng Wu

@uuujingfeng

Postdoc @SimonsInstitute @UCBerkeley; alumnus of @JohnsHopkins @PKU1898; DL theory, opt, and stat learning.

Berkeley, CA Bergabung Ekim 2013
1.1K Mengikuti1.4K Pengikut
Tweet Disematkan
Jingfeng Wu
Jingfeng Wu@uuujingfeng·
sharing a new paper w Peter Bartlett, @jasondeanlee, @ShamKakade6, Bin Yu ppl talking about implicit regularization, but how good is it? We show its surprisingly effective, that GD dominates ridge for all linear regression, w/ more cool stuff on GD vs SGD arxiv.org/abs/2509.17251
Jingfeng Wu tweet media
English
11
33
197
136.5K
Jeremias Sulam
Jeremias Sulam@Jere_je_je·
Belated professional update 🔊 I’ve been promoted to Associate Professor with tenure at @JohnsHopkins @JHUBME. I’m incredibly thankful to my mentors over the past two decades, to my past and current collaborators (and friends!), and immensely proud of my students!
Jeremias Sulam tweet media
English
13
4
61
5.7K
Jingfeng Wu me-retweet
Mahdi Soltanolkotabi
Mahdi Soltanolkotabi@mahdisoltanol·
Fun new theory paper on data reuse: 𝗙𝗨𝗟𝗟-𝗕𝗔𝗧𝗖𝗛 𝗚𝗗 can beat 𝗢𝗡𝗘-𝗣𝗔𝗦𝗦 𝗦𝗚𝗗 by a log-factor in samples. Same single-index model, same data: → 𝗚𝗗 recovers with n ≈ d → 𝗦𝗚𝗗 needs n ≳ d log d arxiv.org/abs/2602.02431
Mahdi Soltanolkotabi tweet media
English
2
13
75
5K
Jingfeng Wu me-retweet
Weijie Su
Weijie Su@weijie444·
It is my tremendous honor and privilege to receive the 2026 COPSS Presidents' Award. Statistics is powerful and will only grow more vital in the AI age. Grateful to my mentors, collaborators, colleagues, and students who made this journey possible.
The Wharton School@Wharton

Congratulations to Prof. Weijie Su (@weijie444) from our Statistics and Data Science Department on being named the recipient of this year's Committee of Presidents of Statistical Societies (@COPSSNews) Presidents' Award: whr.tn/3ZToJu9 The honor is given annually to a young member of the statistical community in recognition of outstanding contributions to the profession of statistics. It's jointly sponsored by five statistical societies: @AmstatNews, @ENAR_ibs, @InstMathStat, @SSC_stat, and @WNAR_ibs.

English
55
24
254
24.8K
Jingfeng Wu me-retweet
The Wharton School
The Wharton School@Wharton·
Congratulations to Prof. Weijie Su (@weijie444) from our Statistics and Data Science Department on being named the recipient of this year's Committee of Presidents of Statistical Societies (@COPSSNews) Presidents' Award: whr.tn/3ZToJu9 The honor is given annually to a young member of the statistical community in recognition of outstanding contributions to the profession of statistics. It's jointly sponsored by five statistical societies: @AmstatNews, @ENAR_ibs, @InstMathStat, @SSC_stat, and @WNAR_ibs.
The Wharton School tweet media
English
1
17
126
28.7K
Jingfeng Wu me-retweet
Data Science Institute
Data Science Institute@DSI_UChicago·
Don't miss the next Statistics and DSI Joint Colloquium! @uuujingfeng, postdoc fellow at @SimonsInstitute at @UCBerkeley, presents 'Towards a Less Conservative Theory of Machine Learning: Unstable Optimization and Implicit Regularization' on Thursday, February 5th at DSI
Data Science Institute tweet media
English
1
3
5
632
Jingfeng Wu me-retweet
TTIC
TTIC@TTIC_Connect·
Thursday, December 11th at 11AM: Talks at TTIC presents Jingfeng Wu (@uuujingfeng) of @SimonsInstitute with a talk titled "A Statistical View on Implicit Regularization: Gradient Descent Dominates Ridge." Please join us in Room 530, 5th floor.
TTIC tweet media
English
1
2
10
768
Jingfeng Wu me-retweet
Spencer Frei
Spencer Frei@sfrei_·
I'm hiring a Student Researcher to work on scaling laws at Google DeepMind! Project is for 16 weeks, starting spring/summer '26, in-person in SF (pic from the amazing office). If you're interested, fill out this form: forms.gle/MsgPfJumTLLobN…
Spencer Frei tweet media
English
19
69
757
73.3K
Jingfeng Wu me-retweet
Sham Kakade
Sham Kakade@ShamKakade6·
1/6 Introducing Seesaw: a principled batch size scheduling algo. Seesaw achieves theoretically optimal serial run time given a fixed compute budget and also matches the performance of cosine annealing at fixed batch size.
Sham Kakade tweet media
English
2
33
247
42.2K
Jingfeng Wu me-retweet
Jeremy Cohen
Jeremy Cohen@deepcohen·
Even with full-batch gradients, DL optimizers defy classical optimization theory, as they operate at the *edge of stability.* With @alex_damian_, we introduce "central flows": a theoretical tool to analyze these dynamics that makes accurate quantitative predictions on real NNs.
English
18
207
1.3K
233.6K
Jingfeng Wu
Jingfeng Wu@uuujingfeng·
sharing a new paper w Peter Bartlett, @jasondeanlee, @ShamKakade6, Bin Yu ppl talking about implicit regularization, but how good is it? We show its surprisingly effective, that GD dominates ridge for all linear regression, w/ more cool stuff on GD vs SGD arxiv.org/abs/2509.17251
Jingfeng Wu tweet media
English
11
33
197
136.5K
Jingfeng Wu me-retweet
Sadhika Malladi
Sadhika Malladi@SadhikaMalladi·
Excited to share that I will be starting as an Assistant Professor in CSE at UCSD (@ucsd_cse) in Fall 2026! I am currently recruiting PhD students who want to bridge theory and practice in deep learning - see here: cs.princeton.edu/~smalladi/recr…
English
38
69
549
86.1K
Jingfeng Wu me-retweet
Jalaj Upadhyay
Jalaj Upadhyay@jalajupadhyay·
I do not think I want to serve as a reviewer of @NeurIPSConf anymore. There are too many checkpoints. Seriously, what the heck is the final justification? Another checkpoint? I review voluntarily because I love my field, but these emails sound more like bullying.
English
12
9
144
17.6K
Jingfeng Wu me-retweet
Mark Schmidt
Mark Schmidt@MarkSchmidtUBC·
Optimization theory focuses on developing algorithms that behave well in the worst case while machine learning only needs an algorithm that works well under one hyper-parameter setting. I think this is a big source of the modern gap between optimization and machine learning.
English
1
3
45
3.5K

Jelajahi

@Jere_je_je @JohnsHopkins @JHUBME @weijie444 @COPSSNews @AmstatNews @ENAR_ibs @InstMathStat