Ruidong Wu

PhD student at @MIT_CSAIL. Prev: @HelixonBio @UofIllinois @Tsinghua_Uni.

[Ruidong Wu @RuidongWu]

Mark your calendars if you want to learn with us—we are live-streaming the lectures all semester:
📆 Days: Tuesdays & Thursdays
⏰ Time: 4:00 - 5:30 PM EDT

[Ruidong Wu @RuidongWu]

Thrilled to join the teaching assistant team for MIT 6.5940: TinyML and Efficient AI Computing this fall! 🛠️⚡Want to tune in? We are streaming our lecture live on YouTube right NOW. Join us here: https://t.co/N9iZxjJPf5

[Shang Yang @Shang_mit]

Excited to co-instruct MIT 6.5940: TinyML and Efficient AI Computing with @songhan_mit this fall!
Looking forward to teaching, learning, and building with an incredible class.

Course site: https://t.co/LMHjNyl0iv

[Shang Yang @Shang_mit]

Excited to co-instruct MIT 6.5940: TinyML and Efficient AI Computing with @songhan_mit this fall!
Looking forward to teaching, learning, and building with an incredible class.

Course site: https://t.co/LMHjNyl0iv

[Ruidong Wu @RuidongWu]

@eric_alcaide Very good question! My personal opinion is that "aligned RMSD" loss of AF3 is actually another chordal distance estimator (similar proof) and will suffer from similar problems. But it is hard to add geodesic terms without explicit frames, and I am still thinking about this.

[Ruidong Wu @RuidongWu]

The widely-used FAPE loss of AlphaFold2 actually suffers from a gradient vanishing problem! By fixing it with geodesic terms with ⭐Frame Aligned Frame Error⭐，we can fix it and improve performance on immune complexes by a large margin!

Presenting at ICML TODAY!😊
(1/4)

[Ruidong Wu @RuidongWu]

@amelie_iska Thank you for the feedback! Already working on it and will be fixed very soon.

[Ruidong Wu @RuidongWu]

@DaveJuergens Thanks David!

[Ruidong Wu @RuidongWu]

@AminSagar1 Thanks for the feedback! I will fix the issue as soon as I get back to my hotel.

[Ruidong Wu @RuidongWu]

@IsomorphicLabs @PeteWirnsberger @riceasphait Really looking forward to it. Is it late to register now?

[Ruidong Wu @RuidongWu]

@json_yim @AndrewC_ML Sorry to hear that, and would love to drop by!

[Ruidong Wu @RuidongWu]

@yisongyue Very insightful work! Already see the poster yesterday and really expect to hear your presentation!

[Ruidong Wu @RuidongWu]

@kb_syx Thanks Yuxuan!

[Ruidong Wu @RuidongWu]

This work can not be done without my amazing friends and collaborators @RuihanGuo2 @RuiWang2017 @luost26 @Ced3LJH @musicsuperman @luoyunan @lqiang67 @_jianpeng and Yue Xu. They have been most helpful!

Again, we are at ICML TODAY session 3 hall C #113. Feel free to drop by! (4/4)

[Ruidong Wu @RuidongWu]

By replacing it with FAFE, which we prove to approximate the geodesic distance instead, we can fix this problem! (3/4)

Paper:
arxiv https://t.co/pcTg9jFdHU
openreview https://t.co/0lH9db7nf7

Code:
https://t.co/d5AGWt744k

[Ruidong Wu @RuidongWu]

@_romain_lopez_ @NYU_Courant Congratulations!

[Ruidong Wu @RuidongWu]

@Haoxiang__Wang @natolambert I use it only because I already paid for 1 year...I guess nowadays we have many better choices.

[Shitong Luo @luost26]

What if your molecules designed by generative models are hard to synthesize?
Project it into a synthesizable chemical space!

That's what our work "Projecting Molecules into Synthesizable Chemical Spaces" does (to appear in ICML 2024).

Paper: https://t.co/TvaG3O7SUG
(1/4)

[Ruidong Wu @RuidongWu]

@abs_WaleedM @sokrypton @ProteinBoston @harvardmed +1😇

[Ruidong Wu @RuidongWu]

@mjstam Congrats! So happy for you!