Latest AI JobsBrowse Jobs
Thinking Machines logo

Research, Mid Training

Thinking MachinesSan FranciscoResearch
Apply on company site
Salary
$350K- $475K USD
Job type
Full-time
Posted
21 Sept 2026
Status
Open · checked 7d ago

About this role

Role Overview

The Research, Mid Training position at Thinking Machines is a full-time individual contributor role based in San Francisco, with remote flexibility available. This position sits within the research department, bridging foundational research and practical engineering to transform base models into reasoning powerhouses. It is an ideal fit for professionals who want to oversee late-stage model training and define core artificial intelligence behaviors.

Key Responsibilities

Own the end-to-end data lifecycle by sourcing, curating, filtering, deduplicating, and synthesizing training-grade datasets that reflect realistic usage.
Design and measure knowledge-increase interventions, including targeted corpora, synthetic rephrasings, and knowledge-dense mixes.
Introduce and evaluate new model behaviors during mid-training to ensure downstream post-training processes become more sample-efficient and stable.
Build and manage automatic filtering and scoring stacks, utilizing quality classifiers and LLM-based judges to raise dataset quality.
Iterate on training recipes by adjusting datasets, stages, annealing schedules, and hyperparameters while tracking downstream impacts.
Define, optimize, and refine meaningful evaluation loops that accurately capture model performance and capabilities.
Debug complex training configurations and investigate unexpected results to build deeper technical insights for future problems.
Scale existing methodologies and explore new training data strategies to measure performance improvements alongside dataset growth.

Requirements

Proficiency in Python programming alongside familiarity with deep learning frameworks such as PyTorch or TensorFlow.
Comfort with debugging distributed training and writing scalable code.
A Bachelor’s degree or equivalent experience in Computer Science, Machine Learning, Physics, Mathematics, or a closely related technical discipline.
Strong theoretical and empirical grounding with clear written communication skills for explaining complex technical ideas.

Nice to Have

A strong grasp of machine learning fundamentals, probability, and statistics to isolate real effects from noise and bugs.
Hands-on experience building or owning large model training datasets, including synthetic data pipelines and real user data curation.
Prior involvement with knowledge injection, continued pre-training, domain adaptation, and measuring model retention.
Experience creating model-based data quality systems, LLM judges, or verification pipelines operating at billions of tokens.
Background in generating synthetic data while understanding failure modes like diversity collapse and contamination.
Past research or engineering contributions focused on alignment, human-AI collaboration, or data-centric artificial intelligence.

Skills and Tools

Python
PyTorch
TensorFlow
Large Language Models (LLMs)
Distributed Training
Data Pipelines
Machine Learning Fundamentals

Salary and Benefits

Annual salary range between $350,000 and $475,000 USD, determined by background, skills, and experience.
Generous health, dental, and vision insurance benefits.
Unlimited paid time off (PTO).
Paid parental leave.
Relocation support provided as needed.
Visa sponsorship available for qualified candidates.

About the Company

Thinking Machines is dedicated to building artificial intelligence systems that extend human will and judgment. The organization is actively training frontier models using Inkling, developing Tinker to allow customized model creation, and designing interfaces to enhance human-AI communication. The company values a human-centered future and encourages equitable hiring practices without discrimination based on protected group statuses.

How to Apply

Interested candidates should visit the official Thinking Machines careers page to submit an application for this position.

Skills and tech stack

PythonPyTorchTensorFlowJAXLLMDeep LearningMachine Learning

Questions candidates ask

What is the location of the Research, Mid Training role?

The position is based in San Francisco, California, with remote options available.

What is the salary range for this position?

The expected annual salary range is between $350,000 and $475,000 USD, depending on background, skills, and experience.

Does the company offer visa sponsorship?

Yes, Thinking Machines sponsors visas and works closely with candidates who are the right fit.

What technical skills are required?

Candidates must be proficient in Python, familiar with deep learning frameworks like PyTorch or TensorFlow, and comfortable debugging distributed training.

What benefits are included with employment?

Benefits include generous health, dental, and vision coverage, unlimited paid time off, paid parental leave, and relocation support.

Ready to apply?

The application is on the official Thinking Machines careers page.

Apply on company site