Remote job
AI Trainer – Computer Science Expert - Graphical Abstract (Fully Remote)
Job details
About this role
Role overview
A human data platform is recruiting computer science specialists to evaluate AI-generated graphical abstracts derived from academic papers. The work supports a project that verifies whether machine-generated visual summaries faithfully represent the underlying research. This is a flexible, ad hoc, fully remote engagement where expert judgement shapes how AI summarises complex technical literature.
Responsibilities
- Review computer science research papers alongside graphical abstracts produced by a large language model. - Compare each abstract against the source paper to confirm it captures the correct methods, findings, and concepts. - Identify factual errors, missing information, or misrepresentations in either the visual layout or the accompanying text. - Apply subject-matter knowledge to judge the structural integrity of each abstract, including how ideas are sequenced and grouped. - Flag cases where technical terminology is used loosely, imprecisely, or in a way that could mislead readers. - Provide written feedback or ratings that can be used to refine AI output quality.
Requirements
- A Master's or PhD in Computer Science, or a closely related discipline. - Demonstrated familiarity with reading and interpreting peer-reviewed computer science literature. - Experience with fact-checking, technical review, or other analytical evaluation work. - Strong attention to detail and the ability to spot subtle inaccuracies in dense material. - Comfort working independently and asynchronously in a remote setting. - Reliable internet access and the ability to complete assessments within given timeframes.
Nice to have
- Prior experience evaluating or annotating outputs for AI training or fine-tuning projects. - A background in scientific publishing, peer review, or journal editing.
Benefits and work setup
- Fully remote, with flexible scheduling suited to ad hoc availability. - Competitive per-task or hourly pay rates. - Onboarding via a short assessment that can typically be completed in around 15 minutes for successful applicants.