wagey.ggwagey.ggv1.0-68eec7a-3-May
Browse Tech JobsCompaniesFeaturesPricingFAQs
Log InGet Started Free
Jobs/Research Scientist Role/apolloresearch - Apollo Research - Research Scientist/Engineer (Science of Scheming)
apolloresearch

apolloresearch - Apollo Research - Research Scientist/Engineer (Science of Scheming)

Remote - London£100k - £200k+ Equity2mo ago
RemoteEMEAArtificial IntelligenceResearch ScientistResearch EngineerApolloPython

Upload My Resume

Drop here or click to browse · Tap to choose · PDF, DOCX, DOC, RTF, TXT

Apply in One Click
Apply in One Click

Requirements

• Fast-paced empirical research: Designing and executing experiments to speed up iteration cycles; excellence in this area. • Conceptual insights about scheming: Familiarity with relevant literature on AI scheming, turning vague concepts into concrete experiment proposals. • Software engineering skills (Python): Strong software engineering abilities for effective execution of work using Python stack. • Intense interest in AI progress: Staying up to date on the latest model releases and continuously tinkering with new workflows; fascination by AI cognition, actively trying to understand how they think. • Experience RL-training LLMs (reinforcement learning): Hands-on experience in training large language models via reinforcement learning techniques. • Strong analytical skills: Background working on fields such as scaling laws in LLMs, statistical physics, dynamical systems, applied statistics; comfortable building mathematical models of empirical phenomena and understanding quantitative aspects related to AI scheming risks evolution with model capability scale.

Responsibilities

• Collaborate with leading AI developers to impact the construction and deployment of capable AI systems through research collaborations. • Study RL dynamics related to reward-seeking behavior, evaluation awareness, and misaligned preferences; design and train model organisms for scaling insights into frontier systems. • Work on developing novel empirical foundations predicting how scheming risks evolve as models scale in capability (Scaling laws of scheming). • Develop new evaluation techniques with the potential to be applied to highly evaluation aware AI models. • Investigate patterns in reasoning processes within frontier AI systems that have not been observed before, focusing on aspects related to cognition and behavioral insights.

Benefits

• Equity options mentioned as part of the benefits package • Paid Time Off (PTO) is included in the compensation plan • Insurance coverage provided to employees • Remote work options offered, allowing for flexibility in working location

Similar Jobs

KyvernaKyverna - Therapeutics - Sr. Clinical Research Scientist2d ago
·Remote
RemoteWWSeniorPharmaceuticalsClinical ResearchBiotechnologyResearch ScientistCRODocumentationPerformance ReviewsClinical DocumentationReportingGCP
relationrxrelationrx - Snr/Principal Machine Learning Scientist – Generative Modelling3d ago
·London, United Kingdom
In OfficeEMEAPrincipalArtificial IntelligenceUtilitiesPrincipalResearch ScientistCSSPython
Sand Tech Holdings LimitedSand Tech Holdings Limited - Senior Test Automation Engineer3d ago
·Remote - EMEA
RemoteEMEASeniorDeveloper ToolsAutomation EngineerSenior Software EngineerTypeScriptPlaywrightGraphQLApolloCursor
Get Started Free

No credit card. Takes 10 seconds.

Privacy·Terms··Contact·FAQ·Wagey on X