Research Scientist, Applied White-Box Methods

We need to output HTML fragment with sections: About the role, Responsibilities? Actually the posting includes "About The Role" and "About You". There's no explicit Responsibilities, Requirements, Qualifications, Skills, Benefits, Pay, Schedule sections. However we have Benefits listed, Compensation (Pay), Hours (Schedule). So we need sections for Benefits, Pay, Schedule if present. Also maybe "About the role" as heading. "About You" could be considered Requirements/Qualifications? The instruction: use

for section headings (e.g. About the role, Responsibilities, Requirements, Qualifications, Skills, Benefits, Pay, Schedule). So we can map content accordingly. We must include a section only when posting has real, specific content for it. So we need sections: About the role (we have that), Benefits (we have bullet list), Pay (compensation range), Schedule (hours full-time 40 hours/week). Possibly also Requirements/Qualifications (we have "About You" which includes qualifications). Could treat that as Requirements or Qualifications. Since we have both "About You" content, we could put under Requirements or Qualifications. The instruction says prefer a few substantive sections over many thin ones. We could combine About You into Requirements (or Qualifications). Let's decide: Use "Requirements" heading for About You content. Or "Qualifications". Which is more appropriate? The About You lists backgrounds, experience, etc. That's qualifications. So use "Qualifications". We'll not duplicate. We also have "About Us" but that's likely boilerplate? It's about the organization, not required. The instruction: remove cookie/navigation/legal/EEO/equal-opportunity/scam-warning boilerplate, duplicated lines, and application instructions. About Us might be considered organizational description; not required but could be kept? The instruction says preserve ALL substantive information; do NOT invent or summarize away detail. About Us is substantive about the organization. However the sections we are to produce are limited to those headings. The instruction says use those headings for section headings. It doesn't forbid other headings? It says use

for section headings (e.g. About the role, Responsibilities, Requirements, Qualifications, Skills, Benefits, Pay, Schedule). It implies we should use those headings for sections. Probably we should only output those sections. The About Us could be omitted as not matching those headings. However the instruction says preserve ALL substantive information; do NOT invent or summarize away detail. If we omit About Us, we lose substantive info about the organization. But maybe the instruction expects we only output those sections; the rest is considered boilerplate? But About Us is not boilerplate; it's descriptive. However the instruction list of possible headings includes "About the role". Not "About Us". So maybe we should not include About Us because it's not one of the allowed headings. The instruction says "use

for section headings (e.g. About the role, Responsibilities, Requirements, Qualifications, Skills, Benefits, Pay, Schedule)." The e.g. suggests those are examples, not exclusive list. But later: "IMPORTANT: include a section ONLY when the posting has real, specific content for it — never emit a heading followed by filler/placeholder text". So we could include any heading we want as long as it's meaningful. But they gave examples. Safer to stick to those examples. Thus we will output sections: About the role, Qualifications, Benefits, Pay, Schedule. We need to preserve all substantive information from those sections. Let's extract content. About The Role: paragraph(s). It says: "As a Research Scientist on the team, you will take ownership of and accelerate the team's research agenda, publish findings broadly, and engage with the AI alignment community. You are encouraged to propose new directions within the team's agenda. You are welcome and encouraged to attend relevant conferences and other community events, and leverage FAR.AI’s existing comprehensive infrastructure for events convening and government relations as you see fit. Beyond FAR.AI, you can work with national AI safety institutes, frontier model developers, and top academics." We'll put that in

under

About the role

. Qualifications (About You): Two backgrounds particularly suited, plus list of qualifications. We need to preserve all detail. Let's extract: "Two backgrounds are particularly well suited to this role: researchers with some experience in more fundamental interpretability who want to evaluate and refine those methods in realistic settings, including agentic coding, long contexts, and realistic threat models; and researchers from evaluations, AI control, red-teaming, or reinforcement learning who have begun working with model internals and want to deepen that work. In both cases, prior hands-on experience with white-box methods and a demonstrated interest in their practical application are preferred. (Note that this role is not well suited to researchers whose primary interest is interpretability as a foundational science.) If you are new to AI Safety research, that is also ok! But please be prepared to explain how your previous research experience (e.g. in applied ML) could be leveraged for our work, and how you are engaging with the field of technical AI safety research today. We are hiring across a range of levels of experience. You may already have some of the following, though not necessarily all. Hands-on experience applying at least one white-box method to a real model (activation explainers, SAEs, steering, attribution, influence functions, probes, or similar), and an informed view of its limitations. A track record in AI safety: a paper, a fellowship project, or substantive public writing. Alumni of programs such as MATS, Astra, Anthropic Fellows, SPAR, Pivotal, LASR or similar programs are especially encouraged to apply. Experience with evaluations, AI control, red-teaming, reinforcement learning, or post-training of LLMs. Experience running large training runs is a plus. The ability to communicate novel methods and results clearly to technical and non-technical audiences. A PhD or several years of research experience in computer science, machine learning, physics, statistics, or a related field. Previous experience in applied ML for other fields: e.g. biology, chemistry, materials science, robotics, etc." We'll put that in

maybe multiple paragraphs but can be one. Benefits: list bullet points. We have: "Health Insurance - 94% of Insurance premium paid by Organization commencing within 1 month after your start date Retirement - 401(k) plan with up to 2% match PTO - 25 days Paid Time Off per year, accrued weekly and up to 10 days of paid sick leave per year Paid Leave - Paid Bereavement, Family, Medical and Pregnancy Disability Leave WFH Stipend & Equipment - Work computer and stipend provided for eligible employees Catered Meals (Berkeley Office Only) - Catered lunches and dinners on workdays at our office(Available only to full-time employees located in the US)" We need to format as

• each. Pay: Compensation range: "$150,000–$250,000/year depending on experience and location, plus work-related travel and equipment expenses. Catered lunch and dinner at our Berkeley office." Also note "Compensation Range: $150K - $250K" at end. We'll put that in

  under

  Pay

  . Schedule: Hours: full-time (40 hours/week). Also location: Berkeley, CA. This role is in-person; we strongly prefer candidates who are in the Bay Area or willing to relocate, and we sponsor visas for in-person employees. We will consider remote arrangements for exceptional candidates. That's schedule/location info. We'll put under

  Schedule

  maybe include location as part. We need to avoid filler like "commensurate with experience". Not present. We must not include job ID etc. We must not repeat job title as heading. We must not include application instructions (like "Application process: interviews..."). We must remove boilerplate like EEO etc. Not present. Now produce HTML fragment. Let's craft:

  About the role

  As a Research Scientist on the team, you will take ownership of and accelerate the team's research agenda, publish findings broadly, and engage with the AI alignment community. You are encouraged to propose new directions within the team's agenda. You are welcome and encouraged to attend relevant conferences and other community events, and leverage FAR.AI’s existing comprehensive infrastructure for events convening and government relations as you see fit. Beyond FAR.AI, you can work with national AI safety institutes, frontier model developers, and top academics.

  Qualifications

  Two backgrounds are particularly well suited to this role: researchers with some experience in more fundamental interpretability who want to evaluate and refine those methods in realistic settings, including agentic coding, long contexts, and realistic threat models; and researchers from evaluations, AI control, red-teaming, or reinforcement learning who have begun working with model internals and want to deepen that work. In both cases, prior hands-on experience with white-box methods and a demonstrated interest in their practical application are preferred. (Note that this role is not well suited to researchers whose primary interest is interpretability as a foundational science.) If you are new to AI Safety research, that is also ok! But please be prepared to explain how your previous research experience (e.g. in applied ML) could be leveraged for our work, and how you are engaging with the field of technical AI safety research today. We are hiring across a range of levels of experience. You may already have some of the following, though not necessarily all. Hands-on experience applying at least one white-box method to a real model (activation explainers, SAEs, steering, attribution, influence functions, probes, or similar), and an informed view of its limitations. A track record in AI safety: a paper, a fellowship project, or substantive public writing. Alumni of programs such as MATS, Astra, Anthropic Fellows, SPAR, Pivotal, LASR or similar programs are especially encouraged to apply. Experience with evaluations, AI control, red-teaming, reinforcement learning, or post-training of LLMs. Experience running large training runs is a plus. The ability to communicate novel methods and results clearly to technical and non-technical audiences. A PhD or several years of research experience in computer science, machine learning, physics, statistics, or a related field. Previous experience in applied ML for other fields: e.g. biology, chemistry, materials science, robotics, etc.

  Benefits

  • Health Insurance - 94% of Insurance premium paid by Organization commencing within 1 month after your start date
  • Retirement - 401(k) plan with up to 2% match
  • PTO - 25 days Paid Time Off per year, accrued weekly and up to 10 days of paid sick leave per year
  • Paid Leave - Paid Bereavement, Family, Medical and Pregnancy Disability Leave
  • WFH Stipend & Equipment - Work computer and stipend provided for eligible employees
  • Catered Meals (Berkeley Office Only) - Catered lunches and dinners on workdays at our office (Available only to full-time employees located in the US)

  Pay

  Compensation: $150,000–$250,000/year depending on experience and location, plus work-related travel and equipment expenses. Catered lunch and dinner at our Berkeley office.

  Schedule

  Hours: full-time (40 hours/week). Location: Berkeley, CA. This role is in-person; we strongly prefer candidates who are in the Bay Area or willing to relocate, and we sponsor visas for in-person employees. We will consider remote arrangements for exceptional candidates