Research Engineer - Scalable Interpretability

Transluce; San Francisco (CA); Sur place; USD 250 000 - 500 000; Plein temps

Résumé du poste

Transluce, a non-profit research lab in San Francisco, seeks scientists and engineers to advance AI oversight tools. You will help develop interpretable assistants and evaluate model behaviors, ensuring industry standards. Ideal candidates have experience in fine-tuning language models, excellent communication skills, and a curiosity for machine learning.

We offer a competitive salary range of $250,000 - $500,000 annually, in a collaborative, in-person work environment, with visa sponsorship available for international talents.

Qualifications

• Experience with fine-tuning language models, designing new architectures, and creating evaluations.
• Reliable results from good experimental design.
• Strong programming skills and ability to navigate trade-offs between speed and maintainability.

Responsabilités

• Help develop and train scalable interpretability assistants.
• Create diverse evaluations that find undesirable model behaviors.
• Scale up training and inference pipelines for large models.

Connaissances

Fine-tuning language models

Experimental design

Strong programming ability

Communication skills

Description du poste

Salary range: $250,000 - $500,000/year + benefits

Description: Transluce is a non‑profit research lab building tools for scalable, end‑to‑end oversight of AI systems. We build world‑class, AI‑backed analysis tools and use these to set industry standards for evaluation. Our tools are integrated with core agent benchmarks like SWE‑bench, while our evaluations are directly underpinning regulation, including our role as EU AI Office’s main evaluation developer for harmful manipulation risks.

About the role:

We are looking for strong scientists and engineers to help advance our vision of scalable end‑to‑end oversight assistants, building on our recent advances such as predictive concept decoders and user model extractors. As part of our highly collaborative team, you will learn and grow quickly, creating technology at the frontier of AI research and with high direct impact.

Core responsibility: Help us develop and train scalable interpretability assistants that can predict and detect unexpected and subtle behaviors from models’ activations. This includes:

• Creating diverse evaluations that range in difficulty. This involves finding naturally occurring interesting and undesirable behaviors exhibited by open‑source models.
• Developing novel architectures and objectives for training interpretability assistants.
• Scaling up the training and inference pipelines to support up to 1T‑scale models.

Qualities of a strong candidate:

• Experience with fine‑tuning language models, designing new architectures, and creating evaluations.
• Reliable results: good experimental design, epistemic self‑awareness and transparency
• Generativeness: coming up with original, productive ideas for unblocking progress
• Curiosity: a desire to understand ML systems and how they work
• Strong programming ability, including navigating trade‑offs between prototyping speed and maintainability
• Strong communication skills, low ego, openness to giving and receiving feedback

We are located in San Francisco and enthusiastic to work together in‑person. We are open to sponsoring international visas.