Founding Scientist, Neural Foundations & Lab AI

Open Culture; Santa Cruz (CA); Remote; USD 155,000 - 215,000; Full time

Benefits offered by this job

Equity

Health benefits

Flexible paid time off

Job summary

Open Culture seeks a computational neuroscience expert to build the model layer that relates recordings from living cultures to public reference data, both in vivo and in vitro, and to decide which experiment runs next.

You will generate with the biology team the data a neural foundation model needs, develop embeddings and metrics, and train the model on spike and field-potential data, evaluated on held-out public data.

Qualifications

• PhD in computational neuroscience, neuroscience, machine learning or bioinformatics.
• Extracellular electrophysiology at scale: spike sorting, population dynamics and the failure modes of HD-MEA or Neuropixels recordings.
• Machine learning on neural time series in PyTorch, trained and evaluated by you.
• Python pipelines other people run, with data versioned as carefully as code.
• Wet-lab literacy and the willingness to plate cells yourself.

Responsibilities

• Relate recordings from our cultures to public in vivo and in vitro reference data, and say how faithfully a culture reproduces real tissue.
• Build the embeddings and metrics that make recordings comparable across labs, platforms and preparations.
• Assemble the corpus with provenance and train the model on spike and field-potential data, evaluated on held-out public data.
• Program the Habitat and electrophysiology agents that run experiments, and monitor them remotely.
• Grow the Scientific Memory Commons, and plate neurons on arrays with the biology team when the work needs it.

Skills

Spike sorting

PyTorch

Python pipelines

Data versioning

Wet-lab literacy

Education

PhD in computational neuroscience or related field

Tools

NWB

DANDI

SpikeLab

Job description

Open Culture seeks a computational neuroscience expert to build the model layer that relates recordings from living cultures to public reference data, both in vivo and in vitro, and to decide which experiment runs next.

You will generate with the biology team the data a neural foundation model needs, develop embeddings and metrics, and train the model on spike and field-potential data, evaluated on held-out public data.