AI Trainer - Computer Science Expert - Graphical Abstract - Fully Remote - (UK)

Role overview

Computer science researchers are invited to join an AI evaluation project as remote reviewers of machine-generated graphical abstracts. The role centres on comparing LLM-produced summaries of academic papers with the original publications to confirm scientific accuracy and visual integrity. It is a flexible, ad hoc engagement that fits around other academic or professional commitments.

Responsibilities

• Review computer science research papers alongside their corresponding AI-generated graphical abstracts
• Identify any discrepancies between the source paper and the abstract in fact, terminology, or emphasis
• Evaluate whether visuals and captions faithfully represent the underlying methodology and findings
• Provide structured written feedback describing issues for the wider evaluation pipeline
• Manage own workload on a task-by-task basis within agreed deadlines

Requirements

• Master's or PhD in Computer Science or a closely related technical discipline
• Track record of engaging with peer-reviewed scientific literature at a critical level
• Strong analytical and fact-checking abilities
• Disciplined attention to detail and consistent evaluative judgement
• Comfortable working independently in a fully remote environment

Nice to have

• Prior exposure to AI training data work, annotation, or model evaluation
• Background in scientific publishing, peer review, or technical copy editing

Benefits and work setup

• Fully remote, work-from-home arrangement (UK-based applicants welcome)
• Ad hoc engagements with flexible hours
• Competitive per-task pay rates
• Fast onboarding once an assessment is passed