Niklas Rindtorff
@Niklas_TR
Followers
3K
Following
13K
Media
91
Statuses
1K
assembling binders @ convexity labs. before: disassembled cancer cells @DKFZ and @broadinstitute
Joined September 2015
It is only called Feynman-Kac when it is from the Feynman-Kac region of France, otherwise it is just sparkling Sequential Monte Carlo.
1
0
2
Catenane-like binding modes when designing cyclic peptides with Boltz-2: The target protein is not cyclic, so not a catenane in sensu stricto. That said, something I haven't seen before.
0
5
49
RT @MSFTResearch: Microsoft researchers achieved a breakthrough in the accuracy of DFT, a method for predicting the properties of molecules….
0
96
0
RT @timsoret: FreeTimeGS: Free Gaussian Primitives at Anytime Anywhere for Dynamic Scene Reconstruction . Paper + demos + code. https://t.co….
0
5
0
Seems like a lot of the pre- vs. post-training compute requirements can be boiled down to "The lower the rate of observing useful training information, the lower the bandwith between nodes can be." . Also great to see @PrimeIntellect getting a shoutout!.
We asked @_sholtodouglas from @AnthropicAI about the costs of RL (Reinforcement Learning) runs. "In @DarioAmodei's essay, he said that RL runs cost only $1M back in December.". "RL is a more naively parallelizable and scalable than pre-training.". "With pre-training, you need
0
0
2
If the largest HPC in 2030 will consume 9 GW (the largest nuclear plants in the world can barely deliver this), what we consider to be noteworthy compute (down to 2 OOM below that number) will still draw as much power as a medium size solar setup / small coal plant produces.
@KonstantinPilz @robi_rahman @james_s48 @ohlennart Read our analysis of the data, alongside comprehensive methods and limitations in our paper. We also provide documentation of the dataset, which answers your frequently asked questions. Paper: Documentation:
0
0
0
RT @SGRodriques: Here's the key graph, comparing performance by ether0 and frontier models across the chemistry tasks we trained + evaluate….
0
5
0
Vance is so good: "The American Space Program, the first program to put a human being on the moon, was built by American citizens. This idea that American citizens don't have the talent to do great things, that you have to import a foreign class of servants, I just reject that."
0
0
5
RT @DrSamuelBHume: This is an incredible progress story. An aggressive type of cancer (Philadelphia chromosome–positive acute lymphoblastic….
0
275
0
RT @BiologyAIDaily: From Likelihood to Fitness: Improving Variant Effect Prediction in Protein and Genome Language Models. １．This study int….
0
19
0
RT @BerkeleyLab: New #AI-ready dataset = a leap forward for drug discovery & energy technologies. @AIatMeta + Berkeley Lab co-led #OMol25….
newscenter.lbl.gov
Scientists will finally be able to simulate the chemistry that drives our bodies, our environment, and our technologies.
0
7
0
What are other targets that have notoriously lousy commercial antibodies right now?.
Antibodies don't work. NLRP3 is by far the most studied inflammasome — yet <30% of commercial antibodies were specific. Only 4 of 14 tested antibodies detected NLRP3 specifically. Antibodies must be validated before use. Full report by @YCharOS1
2
0
3
RT @DrSamuelBHume: The 10 most striking Kaplan-Meier curves (that I have seen) 🧵. 1. Dual GLP1/GIP agonist, Tirzepatide, for prevention of….
0
252
0
Having some fun with @ManifoldMarkets leading up to @AACR this week. (platform is play money only)
1
0
2
