
Thinking about AI safety
AI is analyzing your overall score…
Identifying your key strengths…
Evaluating your skill match against the job requirements…
Assessing your cultural and operational fit
neondatabase
Data Scientist
June 23, 2026 – Present
filtering-for-misalignment
September 3, 2025 – November 19, 2025
What training data should developers filter to reduce risk from misaligned AI? I propose that AI labs filter information about safety measures and strategies for subverting them. This repository helps developers identify such files.
View Projectmisalignment-by-default
March 28, 2025 – June 25, 2025
Model organisms research: do AI goals drift due to "catastrophic forgetting"? Does alignment drift?
View Projectactivation-steering-vs-prompting
March 10, 2025 – Present
Is activation steering more powerful than prompting at mitigating deception in some current reasoning LLMs?
View Projectcoronacoding
March 14, 2020 – April 10, 2020
Coronavacation Coding Club. This is a resource to help people learn to code, specifically over coronavacation.
View Projecttheland
July 7, 2018 – December 8, 2022
A multiplayer video game that I made back in highschool
View ProjectCultural Fit Analysis
The candidate's project portfolio shows a strong inclination towards research-oriented and theoretical aspects of AI, particularly AI safety and alignment. This aligns well with roles requiring deep analytical thinking and a proactive approach to emerging challenges in the field. The diversity of projects, from web development to AI research, indicates a broad technical curiosity. However, the lack of team-based or professional project experience (outside of the current role with no details) makes it difficult to fully assess cultural fit in a collaborative corporate environment.
Soft Skills & Operational Fit
The candidate's numerous personal projects suggest strong self-motivation, curiosity, and a proactive approach to learning and problem-solving. The descriptions of projects like 'filtering-for-misalignment' and 'misalignment-by-default' indicate an ability to tackle complex, abstract problems. However, without psychometric test results or interview data, it is difficult to assess collaboration, stress handling, or communication clarity in a team setting.