Young Investigator, Open Language Models for Biology

The Allen Institute for Artificial Intelligence•Seattle, WA

17d•Onsite

About The Position

Ai2 is seeking a Postdoctoral Young Investigator to work on CellOLMo, a collaboration between Ai2’s open language model researchers and scientists at the Allen Institute. This role involves working at the intersection of machine learning and biology, applying transformer-based models to single-cell data. The researcher will design experiments, work independently, and communicate across disciplines, with a commitment to open science principles by releasing code, data, and model weights. CellOLMo investigates how open language models can be used with single-cell brain data to answer biological questions, focusing initially on the SEA-AD Alzheimer’s disease atlas. A key research question is whether grounding models in biological language and knowledge improves performance over models trained solely on molecular data, and whether these improvements persist at higher levels of representation (brain regions, patients, pathology, disease progression). The investigator will work with the project PI at Ai2 and collaborate with domain scientists at the Allen Institute for Brain Health, with access to Ai2’s open language models, post-training infrastructure, and GPU compute.

Requirements

A completed PhD, or a PhD expected within one year, in machine learning, computational biology, computational neuroscience, physics, or a related quantitative field.
Hands-on experience training or fine-tuning transformer-based language models.
Strong Python and PyTorch skills.
Experience writing reliable, reproducible research code.
The ability to carry out independent research on open-ended problems.
The ability to communicate clearly with both machine-learning researchers and biological scientists.

Nice To Haves

Single-cell genomics, including scRNA-seq or snRNA-seq.
Tools such as scanpy, AnnData, or the scVI ecosystem.
Multimodal models, instruction tuning, or language-model post-training.
Neurodegeneration, brain atlases, or spatial transcriptomics.
Public releases of research code, datasets, or model weights.

Responsibilities

Develop a multimodal language model that combines single-cell gene-expression data with text, using an open model such as OLMo.
Reproduce and evaluate relevant approaches, including scGPT, Geneformer, CellWhisperer, and C2S-Scale.
Adapt and improve these approaches for brain and neurodegeneration data.
Design controlled experiments to test whether language grounding improves performance over cell-only and text-free baselines.
Develop representations at the brain-region and donor levels that incorporate pathology and disease progression.
Evaluate models on held-out donors using measures such as cell-type and marker recovery, agreement with known patterns of regional vulnerability, and expert review of plain-language answers.
Work with Allen Institute scientists to interpret results and identify predictions suitable for experimental follow-up.
Release data, code, and model weights, and prepare the results for publication.

Benefits

Medical, dental, vision, and an employee assistance program for team members and their families.
Health savings account plan.
Healthcare reimbursement arrangement plan.
Health care and dependent care flexible spending account plans.
Company’s 401k plan.
$125 per month to assist with commuting or internet expenses.
$200 per month for fitness and wellbeing expenses.
Up to ten sick days per year.
Up to seven personal days per year.
Up to 20 vacation days per year.
Twelve paid holidays throughout the calendar year.
Annual bonuses.