Chaitanya K. Joshi
@chaitjo
PhD student @Cambridge_CL. Research Intern @AIatMeta and @PrescientDesign. Interested in Deep Learning for biomolecule design.
Are we hitting a wall with scaling molecular foundation models? While LLMs transfer knowledge across domains, molecular models often don't - We may need unified representations for true generalization. Some thoughts on what I think is one of the most interesting open Qs atm:
Exciting new protein design technical reports keep dropping! Wouldn’t it be funny if Latent-X was **not** an all-atom latent diffusion model? 🤣 From this fig. it seems likely that the arcitecture is really fast, likely w/out equivariance (though RFdiff is known to be slow..)
Introducing Latent-X — our all-atom frontier AI model for protein binder design. State-of-the-art lab performance, widely accessible via the Latent Labs Platform. Free tier: platform.latentlabs.com Blog: latentlabs.com/latent-x/ Technical report: tinyurl.com/latent-X
Aithyra Opening Symposium "AI for Life Science" with Nobel Laureate Frances Arnold as keynote speaker in addition to an outstanding line up of speakers on a variety of topics across biological scales and data modalities. Registration is now open at lnkd.in/dkJJCmqx
Yes I think of rational design of materials at the atomistic scale as a keystone problem: Once you unlock it the possibilities are endless and will set off a cascade of breakthroughs throughout the sciences and engineering.
30% of the global economy depends on advanced materials. Yet we still discover them the way we did 200 years ago: mix, heat, test, repeat. At Orbital, we use AI to design materials and because we’re fully verticalized, turn them into deployable hardware. Our CEO Jonathan…
Congratulations to the lead authors and @ShuiwangJi for seeing this survey through to publication at a historic venue! Its not easy to firstly put together such a comprehensive survey, and to publish it. Its years of work - And the community really appreciates the resource! 🙏
Our 500+ page AI4Science paper is finally published: Artificial Intelligence for Science in Quantum, Atomistic, and Continuum Systems. Foundations and Trends® in Machine Learning, Vol. 18, No. 4, 385–912, 2025 nowpublishers.com/article/Detail…
There is an opening for a University Assistant Professor in Machine Learning to be based at @CambridgeMLG jobs.cam.ac.uk/job/49361/ Apply!
Our 500+ page AI4Science paper is finally published: Artificial Intelligence for Science in Quantum, Atomistic, and Continuum Systems. Foundations and Trends® in Machine Learning, Vol. 18, No. 4, 385–912, 2025 nowpublishers.com/article/Detail…
In the long run, I think VCs should encourage sharing, and taking advantage of "replica exchange". Where academia (high temp chain) is where folks try crazy ideas (rfdiffusion, bindcraft, etc), that are further optimized by industry (low temp chain) with far more resources. (2/2)
Come join us next week! London has many "biomolecular modeling" companies and academic labs, yet there's no stand-out in-person forum, where everyone can discuss the latest advancements together - let that be in modeling of proteins, RNA, and/or all biomolecules of life!…
@astar_gis is finally old enough to rent a car! Join our big 25th anniversary conference in Singapore Sep 11-12 to learn how genomics is changing, and how it's changing health.
🎯Come Tackle The 3 Big Questions in Genomics with Us in Singapore This September! 1) Can we have better insights into our health by reading our DNA more accurately? 2) What are the new innovations in genome technology that will change the way we understand human health and…
Somehow this seems less controversial now 😁
A lot of people are skeptical that self-training can work. But the story of Ramanujan shows that once a certain threshold of intelligence is crossed, pure self-training in mathematics is possible even without an external reward signal provided by a proof checker.
What happened to pathology AI companies? owlposting.com/p/what-happene… 4k words, 19 minutes reading time some longer thoughts: i used to hear a lot about digital pathology companies like pathai and paige. i remember listening to podcasts about them and seeing their crazy raises…
The Bitter lesson does not say to not bother with methods research. It says to not bother with methods that are handcrafted datapoints in disguise.
Top 3 AI researchers of all time (in no order) 1. Ilya 2. Noam shazeer 3. Alec Radford Wdyt?
Our IMO gold model is not just an "experimental reasoning" model. It is way more general purpose than anyone would have expected. This general deep think model is going to be shipped so stay tuned! 🔥
So happy to see this incredible achievement. Huge congrats to @lmthang, @quocleix, @YiTayML and the IMO team on the result. This was a great collaboration across teams to build a general Gemini DeepThink model that can also get gold at IMO.
From Silver to Gold🥇 Our advanced version of Gemini Deep Think has achieved the gold-medal standard at the 2025 International Mathematical Olympiad. It perfectly solved 5/6 problems in natural language with no expert intervention and that too within 4.5 hours!
An advanced version of Gemini with Deep Think has officially achieved gold medal-level performance at the International Mathematical Olympiad. 🥇 It solved 5️⃣ out of 6️⃣ exceptionally difficult problems, involving algebra, combinatorics, geometry and number theory. Here’s how 🧵
Very excited to share that an advanced version of Gemini Deep Think is the first to have achieved gold-medal level in the International Mathematical Olympiad! 🏆, solving five out of six problems perfectly, as verified by the IMO organizers! It’s been a wild run to lead this…
Super thrilled to share that our AI has has now reached silver medalist level in Math at #imo2024 (1 point away from 🥇)! Since Jan, we now not only have a much stronger version of #AlphaGeometry, but also an entirely new system called #AlphaProof, capable of solving many more…
Official results are in - Gemini achieved gold-medal level in the International Mathematical Olympiad! 🏆 An advanced version was able to solve 5 out of 6 problems. Incredible progress - huge congrats to @lmthang and the team! deepmind.google/discover/blog/…
💥 Excited to introduce Bacformer 🦠 - the first foundation model for bacterial genomics. Bacformer represents genomes as sequences of ordered proteins, learning the “grammar” of how genes are arranged, interact and evolve. Preprint 📝: biorxiv.org/content/10.110… 🧵 1/n
my emotional biotech portfolio: rowansci.com noetik.ai axi.om popvax.com convoke.bio formation.bio markov.bio futurehouse.org newlimit.com leash.bio
Happy to present the first complete genome of an Indian individual biorxiv.org/content/10.110…. Joint effort by @astar_gis ,FER, and partners, with support from the Singapore NPM programme! Joint effort with JJ Liu Lab! @msprasad693 and @JosipaLipovac as co-first authors.