You'll work with Python and ML pipelines to advance multilingual large language models at Cohere, a security-focused enterprise AI company building foundation models and end-to-end products. As a Member of Technical Staff on the Multilingual team, you'll design scalable solutions to improve multilingual LLM performance across various NLP tasks, publish research, and mentor junior team members.

OpenAI★4.6 (9k)· London

You'll design and build scaled backend and data systems to detect, track, and enforce against harm and abuse on OpenAI's platforms, leveraging the company's advanced AI technologies to automate harm detection and reasoning. Working with trust and safety experts, policy, legal, and investigations teams, you'll collaborate across technical and non-technical roles to combat adversarial actors exploiting OpenAI's technology, requiring 5+ years backend/data systems experience and 2+ years in trust and safety operations.

Anthropic★5.0 (1)· London

You'll conduct empirical ML research using Python to design and run safety experiments on large language models, including adversarial testing, reinforcement learning evaluations, and jailbreak analysis. The Alignment Science team at Anthropic focuses on AI control and alignment stress-testing to understand failure modes in advanced AI systems. You'll need significant software or research engineering experience and familiarity with AI safety research, collaborating across Interpretability, Fine-Tuning, and Red Team groups.