
PhD student at @MIT_CSAIL. Prev: @HelixonBio @UofIllinois @Tsinghua_Uni.
Joined March 2019
- Tweets38
- Following576
- Followers235
- Likes300
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: youtube.com/watch?v=iAZxjWt7…
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: efficientml.ai
2
4
311
Ruidong Wu retweeted
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: efficientml.ai
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)
By replacing it with FAFE, which we prove to approximate the geodesic distance instead, we can fix this problem! (3/4)
Paper:
arxiv arxiv.org/abs/2407.01649
openreview openreview.net/forum?id=Lhb3…
Code:
github.com/mooninrain/FAFE
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 retweeted
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: arxiv.org/abs/2406.04628
(1/4)
Ruidong Wu retweeted
Announcing AlphaFold 3: our state-of-the-art AI model for predicting the structure and interactions of all life’s molecules. 🧬
Here’s how we built it with @IsomorphicLabs and what it means for biology. 🧵 dpmd.ai/3URDiNo
Ruidong Wu retweeted
🥁 Llama3 is out 🥁
8B and 70B models available today.
8k context length.
Trained with 15 trillion tokens on a custom-built 24k GPU cluster.
Great performance on various benchmarks, with Llam3-8B doing better than Llama2-70B in some cases.
More versions are coming over the next few months.
llama.meta.com/llama3/
Ruidong Wu retweeted
Our method for molecular linker design with diffusion models (DiffLinker) got published in @NatMachIntell 🥳
nature.com/articles/s42256-0…
Thanks a lot to our amazing coauthors @HannesStaerk Clément Vignac @rneschneuing @vgsatorras @pafrossard @wellingmax @mmbronstein @befcorreia
Ruidong Wu retweeted
This'll be fun 🧬
Next Tuesday, April 16th @ 4 pm ET we're very excited to have @BrianHie @exnx present Evo: sequence modeling and design from molecular to genome scale
Read the preprint here: biorxiv.org/content/10.1101/…
Sign up on our website to be added to the mailing list to receive Zoom links!
Ruidong Wu retweeted
Michel Talagrand has been awarded the Abel Prize, one of the highest honors in mathematics, for applying tools from high-dimensional geometry to complex probability problems. @jordanacep reports: quantamagazine.org/michel-ta…
Ruidong Wu retweeted
So proud about being able to leverage language models to fold proteins from just their primary sequence. Presenting Omegafold, in collab with @peng_illinois and @HelixonBio :
biorxiv.org/content/10.1101/…
Our OmegaFold paper is now released.
helixon.s3.amazonaws.com/ome…
Thanks so much for my great coauthors, including @peng_illinois @musicsuperman @RuiWang2017 @luost26 @XiwenZhang20 @ChenpengSu at @HelixonBio. It is really a privilege to work with you!
We are thrilled to share the manuscript of OmegaFold.
It is really exciting to a protein's primary sequence is sufficient to predict its folded form. We believe that this conceptual advance will enable a multitude of applications in protein research.
helixon.s3.amazonaws.com/ome…
Ruidong Wu retweeted
I just officially started my appointment @PrincetonCS and couldn't be more excited! Current status: wandering around Princeton figuring out how to buy GPUs #newPI – at Princeton, NJ
Ruidong Wu retweeted
Protein structure can be predicted from a single sequence alone with high accuracy.
@HelixonBio team have developed OmegaFold, achieving performance similar to RF and AF2's MSA versions. Only a single sequence is given as input. 1/5






















