This page was automatically translated and may contain errors. View in English.
跨越障碍

English Speaking Generalist

Crossing Hurdles

Remote · 合同

抢先申请

经验
任何
薪水
职位空缺
1
发布
5天前
工作模式
在家办公
恢复
需要申请

职位描述

Role Overview

This position involves evaluating large language models through the creation and comparison of various prompts while working remotely. Candidates will interact with AI platforms, providing in-depth verbal feedback in American English and ensuring adherence to evaluation standards.

Key Responsibilities

  • Develop diverse prompts on a range of subjects to effectively challenge AI language models.
  • Conduct parallel testing by submitting identical prompts to ChatGPT and Claude, while recording both screen activity and audio commentary.
  • Analyze and articulate differences between model outputs, providing real-time verbal evaluations.
  • Deliver clear verbal reasoning and preferences using fluent American English.
  • Interpret project documentation to maintain consistent evaluation criteria and ensure precise task execution.
  • Collaborate within project frameworks to aid the ongoing improvement of AI models.

Candidate Requirements

  • Exceptional command of spoken and written American English with clear enunciation.
  • Experience interacting with large language models like ChatGPT, Claude, or equivalents.
  • Knowledge or familiarity with AI evaluation methodologies.
  • Comfortable performing extended screen and audio recording sessions.
  • Prior exposure to data annotation, AI model training, quality assurance, content analysis, or research preferred.
  • Possess a robust technical setup including a dependable computer and stable internet connection.

Application Details

  • Application is straightforward and begins with submission.
  • Successful candidates will receive follow-up emails outlining subsequent steps.
  • An interview and resume assessment will be part of the selection process.

如果您希望收到回复,请留下您的信息——我们不会将您的信息用于其他用途。

点击浏览拖放,或 粘贴 截图

PNG、JPG、GIF、MP4、WebM、MOV 格式 · 每个文件最大 20MB · 最多 5 个文件

🤖
在线·即时人工智能帮助