AI Trainer - Neuroscience (Fully Remote- USA)

Role overview This fully remote opportunity calls on neuroscience specialists to assess graphical abstracts generated by AI against the source papers they are meant to summarise. The role feeds into a verification project designed to maintain scientific rigour as large language models take on more technical summarisation tasks. It operates as a flexible, ad hoc engagement suited to academics who want to contribute expertise outside their primary role.

Responsibilities - Compare peer-reviewed neuroscience papers with graphical abstracts drafted by a large language model. - Pinpoint factual inaccuracies, conceptual misstatements, and unsupported visual claims. - Assess whether the abstract's depiction of methodology, data, and conclusions aligns with the original study. - Provide detailed, structured feedback highlighting specific issues for downstream model improvement. - Successfully complete a paid qualifying task involving a single review-and-abstract pair.

Requirements - Master's or PhD in neuroscience or a related scientific discipline. - Comprehensive understanding of contemporary neuroscience research methods and findings. - Demonstrated capability for rigorous, detail-oriented scientific review. - Prior experience with fact verification or critical evaluation of technical content. - Self-directed work style suitable for a remote, project-based arrangement.

Nice to have - Familiarity with AI training pipelines, data annotation, or model evaluation projects. - Involvement in scientific publishing, peer review, or journal editorial processes.

Benefits and work setup - Fully remote with flexible hours for US-based participants. - Hourly compensation beginning at $40+, with the initial assessment also paid.