RESEARCH
Notes From the Lab
Research, Releases, & Perspectives
The latest on Omnii: new research, model releases, partnerships, and perspectives from the Radical Numerics team.
LATEST FROM THE LAB
Introducing Omnii-1: a leap toward general biological intelligence
Introducing Omnii-1, a multimodal model for biology, with eight applications across genome regulation, biomolecule design, health, biosecurity and rare earth separation.
ALSO FEATURED
Radical Numerics creates first AI-generated phages to reach human patient with Omnii-1
READ ARTICLE
Generative multiomic modeling with Omnii-1
READ ARTICLE
Applying Omnii-1 across the genetic architecture of human disease
READ ARTICLE
Biodefense with Omnii-1, through API and edge deployment
READ ARTICLE
ARCHIVE
All Updates
Introducing Omnii-1: a leap toward general biological intelligence
Introducing Omnii-1, a multimodal model for biology, with eight applications across genome regulation, biomolecule design, health, biosecurity and rare earth separation.
Omnii-1 in the Loop: Accelerating Antibody Design with Experimental Feedback
We mid-trained Omnii-1, our multimodal genome language model, for lab-in-the-loop antibody binder design.
Biodefense with Omnii-1, through API and edge deployment
We've put Omnii-1 behind a restricted-access biodefense API and onto a laptop, detecting divergent pathogen sequences in proteins and metagenomic reads that alignment-based tools miss.
Omnii-1 for enhanced multi-cancer diagnostics
We fine-tuned Omnii-1 to predict a proprietary cell-free DNA biomarker it had never seen before and used the resulting model to enhance the sensitivity of a cell-free DNA test for early-stage cancer.
Applying Omnii-1 across the genetic architecture of human disease
We apply Omnii-1 across the genetic architecture of human disease, from closing the diagnosis gap in rare disease to nominating candidate variants behind Parkinson's GWAS signals.
Generative multiomic modeling with Omnii-1
Omnii-1 is a multimodal genome language model. It reconstructs held-out tracks, identifies the variants that change expression, splicing, and accessibility, and links enhancers to their target genes.
Radical Numerics creates first AI-generated phages to reach human patient with Omnii-1
We trained Omnii-1 to generate viable phages, prioritize variants, and predict phage-host interactions. In a world first, Omnii-1-designed phages were administered to a human patient.
Finding proteins that separate rare earth metals, with Omnii-1
With Lawrence Livermore National Laboratory, we're designing proteins that separate rare earths. Omnii-1 predicts their selectivity better than every model we tested.
Radical Numerics Welcomes Jill Hruby and Paul Friedrichs as Biodefense Advisors
Jill Hruby and Paul Friedrichs join Radical Numerics as biodefense advisors, bringing deep experience in national security and biological risk reduction.
Tri Dao joins Radical Numerics as scientific advisor
Tri Dao, Co-Founder and Chief Scientist of Together AI and creator of FlashAttention and Mamba, joins Radical Numerics as scientific advisor on ML systems, optimization, and scaling.
Text-to-image generative modeling without language pretraining
Halcyon is a family of early-fusion multimodal autoregressive models pretrained from scratch on image-text pairs, with no internet-scale language pretraining.
Scaling data mixing for language and DNA pretraining
We scale data mixing for language and genome models: thousands of sources derived from document behavior, and a Pareto frontier of mixtures over them.
Christine Parthemore on joining Radical Numerics as Head of Biodefense
Christine Parthemore joins Radical Numerics as Head of Biodefense, bringing two decades of experience at the U.S. Department of Defense and the Council on Strategic Risks.
Jacob Rinaldi joins Radical Numerics as Head of Scientific Strategy
Jacob Rinaldi joins Radical Numerics as Head of Scientific Strategy, bringing scientific depth from Genentech, Recursion, and Noetik to our work on general biological intelligence.
Genome language models can design cancer vaccines
We post-trained Omnii, our next-generation genome language model, for end-to-end cancer vaccine design, from neoantigen selection to the final mRNA.
Radical Numerics raises $50M seed to build general biological intelligence
We've raised $50M to build general biological intelligence, and we're previewing Omnii, our next-generation genome language model for both human health and biodefense.
NVFP4 pretraining: from theory to implementation (Part 1)
NVIDIA's NVFP4 pretraining recipe: floating point fundamentals, the evolution of mixed-precision training, and the techniques that enable stable NVFP4 training.
NVFP4 pretraining: systems optimizations (Part 2)
A deep dive into TransformerEngine's implementation of the NVFP4 recipe: nvfp4 data flow, custom kernels, and systems optimizations for performant FP4 training.
Phalanx: hardware-aligned sliding-window recurrences
Phalanx: a drop-in replacement for sliding window attention. Faster and better quality. Hardware- and numerics-aware design of new layers that push the efficiency-quality frontier
RND1: simple, scalable AR-to-diffusion conversion
Introducing RND1-30B, the largest open-source diffusion language model, trained via AR-to-Diffusion conversion with a simple, scalable recipe.
How AI could generate new life forms
Eric's TED talk on next-generation AI models for generative genomics.
Partner with Radical Numerics
We work with researchers, companies, and institutions to apply Omnii to challenges across disease, therapeutics, and biosecurity. Stay in touch with us below.