R

Radical Numerics

Develops multimodal foundation models for biology

Member of Technical Staff - Post-Training

Applied Machine LearningConfirmed live in the last 24 hours
Full-timeNo salary listedIn-person

About the Job

Requirements

  • A strong track record in machine learning research or engineering, particularly in frontier-model training, post-training, alignment, evaluation, data quality, or related areas.
  • Proficiency in building production-quality software and research infrastructure, ideally using Python and PyTorch, and comfort debugging large-scale training workflows.
  • Ability to design careful experiments, interpret ambiguous results, and distinguish real effects from artifacts, bugs, or benchmark overfitting.
  • Excellent written and verbal communication skills, including the ability to explain technical findings clearly to research, engineering, and scientific collaborators.
  • Curiosity, rigor, and an iterative approach to improving systems through hypotheses, experiments, and insight.

Responsibilities

  • Develop and tune post-training stages, datasets, reward signals, and hyperparameters for biological world models, studying how data mixtures, objective design, curriculum, and training schedules affect model behavior.
  • Collaborate with the science team to develop and refine evaluation suites for biological reasoning, scientific usefulness, long-context behavior, robustness, and model reliability, and identify when benchmarks are no longer informative and should be replaced.
  • Investigate failure modes in training runs and model outputs, distinguish signal from noise, and trace problems to data, optimization, evaluation design, or systems issues.
  • Explore preference- and feedback-driven learning methods, including preference modeling, reward modeling, synthetic feedback, and related post-training approaches to improve model responses to scientific tasks and constraints.
  • Help define, curate, or generate high-quality post-training datasets, including expert-informed data, synthetic data, and task-specific examples grounded in biological workflows.
  • Measure how post-training performance changes with dataset size, recipe complexity, compute budget, and model family, and use results to guide scaling and future research directions.
  • Work with colleagues in training systems, architecture, and biology-facing research to ground post-training methods in large-scale experimentation and downstream scientific use.

Desired Qualifications

  • Experience with reinforcement learning from human feedback, reinforcement learning from AI feedback, preference optimization, reward modeling, rejection sampling, or other post-training methods for large models.
  • Experience designing or operating evaluation frameworks for model quality, reliability, safety, or scientific task performance.
  • Familiarity with synthetic data generation, annotation workflows, or expert-in-the-loop data collection.
  • A background in applied mathematics, systems, computational biology, or another quantitative scientific field.
  • Contributions to open-source machine learning systems, model tooling, or research infrastructure.

About the Company

Radical Numerics

Develops multimodal foundation models for biology

Radical Numerics is an AI research lab that builds multimodal foundation models for biology to enable generative genomics for medicine and biodefense. Its Evo and Evo 2 models are open biological sequence models trained on millions of genomes, while Omnii is a genome language model with a two-million-token context, multimodality, and alignment that is being piloted for early cancer detection and biosurveillance. The company differentiates itself by focusing on large, multimodal biology models and dual-use health and biodefense applications, often partnering with industry and government labs. Its goal is to scale compute and architecture research to advance practical deployments in human health and biodefense.

Company size
11-50
Company stage
Seed
Total funding
$50M
Headquarters
San Francisco, CA
Founded
2025

Recently Posted Jobs

Explore more opportunities at Radical Numerics

See more jobs