https://chriskambimbi.github.io Chris Kambimbi is a Research Engineer at LibrAI, investigating vulnerabilities in large language models. Writing on AI safety, red teaming, and AI alignment. en https://chriskambimbi.github.io/forty-open-problems-in-ai-safety/ https://chriskambimbi.github.io/forty-open-problems-in-ai-safety/ Wed, 16 Sep 2026 00:00:00 GMT Chris Kambimbi A 2026 update to Concrete Problems in AI Safety: forty open research problems, and the funders committed to them. https://chriskambimbi.github.io/continuous-red-teaming/ https://chriskambimbi.github.io/continuous-red-teaming/ Tue, 15 Sep 2026 00:00:00 GMT Chris Kambimbi Why safety testing must operate at the speed of AI agents. https://chriskambimbi.github.io/deepseek-v32/ https://chriskambimbi.github.io/deepseek-v32/ Thu, 04 Dec 2025 00:00:00 GMT Chris Kambimbi How an open-source release closed the gap with frontier models: sparse attention, large-scale RL post-training, and 1,827 synthesized agentic task environments. https://chriskambimbi.github.io/rejection/ https://chriskambimbi.github.io/rejection/ Tue, 29 Oct 2024 00:00:00 GMT Chris Kambimbi A short poem about a cold letter, and the resolve that follows it.