Research Engineer (Neo Research, Contract)

Neo Research is an independent frontier safety research and evaluations organisation, based in Singapore. We focus on loss-of-control and harmful manipulation risks in frontier AI models.

What you'll do

• Run safety evaluations (e.g. CyBench, SHADE-Arena) against target models.
• Perform open-ended research on models and/or eliciting dangerous behaviours (e.g. scheming, sandbagging, self-replication).
• Draft sections of safety reports documenting methodology, elicitation, and findings.

What we are looking for

• Hands-on experience running or creating LLM safety evaluations.
• Comfortable with Python and agentic evaluation frameworks (Inspect or similar).
• Rigorous about elicitation methodology: scaffolding, tool access, sampling, and knowing where evals fail.
• Clear technical writing.

Good to have

• Dangerous capability evaluations experience.
• Familiarity with frontier lab safety reports.
• Reading and writing in Mandarin.

Concrete details

• 3–6 hour paid work test to start.
• 1–3 months contract, with potential for full-time transition.
• Remote-friendly.