CareerPlanSign in

Research Engineer - Model Architectures

San Francisco💼 Full-time🗓 2026-03-17 → 2026-09-26

Core

Designing and rigorously testing novel model architectures and training methodologies to improve core modeling capabilities and address fundamental bottlenecks in contemporary models.

Role type

Research Engineer - Model Architectures

Builds

Next-generation AI models with improved efficiency and capabilities

Domain

Artificial Intelligence / Machine Learning

Deliverable

production ML models

Required skills

PyTorch, Python, experimental methodology, ablation testing, GPU hardware optimization, reinforcement learning, control theory, signal processing, RAG/retrieval systems, dynamic/adaptive computation, alternative credit assignment approaches

Preferred skills

Long-term memory systems, ability to prototype quickly, unique perspective on 'crazy' ideas, jumping into large codebases

Technologies

PyTorch, Python

Responsibilities

Design novel model architectures, test training methodologies, collaborate with pre-training team, run rigorous ablations and hypothesis testing, optimize for GPU hardware efficiency

Seniority

Mid-Senior, hands-on IC

Sourced via ashby · Listed on CareerPlan, which tracks 70,000+ jobs from 20+ sources.