Kyle Swanson

Excited to share @tutorintel's Data Factory 1, a 100 robot semi-humanoid research farm and the largest robot data factory in the United States. Our first embodiment “Cassie” is deployed at industrial scale across the supply chain. We built DF1 to bootstrap fleet-scale learning for our "Sonny" industrial semi-humanoid embodiment, powered by our first end-to-end robot foundation model Ti0.

Thank you @alanchengsf, Maria Chiriac, @james_y_zou, and all of our other colleagues at Merck for a great project and a really unique opportunity to apply SyntheMol-RL to real-world drug discovery programs! 5/5

We share our results and learnings on two internal programs at Merck: a human therapeutic program and an antibacterial program. We were excited to see promising preliminary results on both projects. We also discuss challenges and share takeaways for future work. 4/5

Following our publication of SyntheMol-RL last week, I'm excited to share the results and learnings from an extensive collaboration with @Merck applying SyntheMol-RL to two of their internal drug discovery pipelines. 1/5 swansonkyle.com/blog/synthemol…

>700K people die every year due to S. aureus infection. SyntheMol-RL created a new molecule that stops drug-resistant S. aureus MRSA—validated in mice studies. AI accelerating drug discovery; published today!👇 Great collab w/ @ItsJonStokes' lab led by @KyleWSwanson Gary Liu

SyntheMol-RL is fully open source for everyone to use. Thank you Gary Liu, Denise Catacutan, @ItsJonStokes, @james_y_zou, and our collaborators for a fun project! 6/6 github.com/swansonk14/Syn…

We tested one particularly potent molecule, which we call synthecin, in a mouse skin infection model of methicillin-resistant S. aureus (MRSA) and found that it was safe and highly effective at treating the infection. 5/6

SyntheMol-RL has now been published! SyntheMol-RL is a reinforcement learning model for synthesizable small molecule drug design. We used it to design antibiotic candidates for the bacteria S. aureus with hits validated in vitro and in vivo in mice. 1/6 link.springer.com/article/10.103…

Today we’re launching Phylo, a research lab studying agentic biology, backed by a $13.5M seed round co-led by @a16z and @MenloVentures / Anthology Fund @AnthropicAI. We’re also introducing a research preview of Biomni Lab, the first Integrated Biology Environment (IBE), where we’re imagining a new way biologists work. Biomni Lab uses agents to orchestrate hundreds of biological databases, software tools, molecular AI models, expert workflows, and even external research services in one workspace, supporting research end-to-end from question to experiment to result. Agents handle the mechanics, while you define the question, then review, steer, and decide. Scientists end up spending more time on science: asking questions, understanding mechanisms, and eliminating diseases. Phylo (@phylo_bio) is a spin-out of @ProjectBiomni, where we will maintain the open-source community and push open-science research. I’m grateful to continue building with my co-founders @YuanhaoQ @jure @lecong and the dream founding team @serena2z @TianweiShe @huangzixin20151 @gm2123 @margaretwhua @malayhgandhi. We’re also fortunate to be advised by leading scientists @zhangf, Carolyn Bertozzi, and @fabian_theis, and supported by an amazing group of investors including @JorgeCondeBio @zakdoric Matt Kraning @ZettaVentures @dreidco @conviction @saranormous @svangel @valkyrie_vc and others. Biomni Lab is available for free today: biomni.phylo.bio Learn more in our launch post: phylo.bio/blog/company-f… We are also hosting launch events - join us at South San Francisco: luma.com/n8k8qb0n Virtual: luma.com/l5ryjaij We’re also hiring! phylo.bio/careers

How to get AI to make discoveries on open scientific problems? Most methods just improve the prompt with more attempts. But the AI itself doesn't improve. With test-time training, AI can continue to learn on the problem it’s trying to solve: test-time-training.github.io/discover.pdf

A huge thank you to Ruhi Sayana, Kate Callon, Jennifer Xu, Jonathan Deutsch, Steven Chu, @james_y_zou, @jjanetzko, and Robin Shivnaraine for the really fun collaboration!

I'm excited to share SyntheFluor-RL, a generative AI model for designing easily synthesizable small molecule fluorescent dyes. Please see below for more details. Paper: arxiv.org/abs/2601.07145 Code: github.com/swansonk14/Syn…

Multiple compounds showed different levels of fluorescence emission intensity, including one especially bright compound in the UV range, validating the efficacy of SyntheFluor-RL.

We applied SyntheFluor-RL to generate 11,590 candidate molecules, which we filtered to a set of 19 structures predicted to possess dye-like properties. We successfully synthesized 14 of those compounds and experimentally tested 13 of them.

Small molecule fluorescent dyes are critical for many bioimaging experiments. However, despite years of research, there are few distinct small molecule dye scaffolds, limiting our options when selecting dyes for an experiment. SyntheFluor-RL enables the rapid design of new dyes.

SyntheFluor-RL extends our previous model, SyntheMol-RL, by introducing new chemical reactions along with property predictors and multi-parameter optimization tailored to fluorescent dye design (e.g., emission wavelength and photoluminescence quantum yield).