Shizuyasu Takeyama

Key Skills

Software

Other

Remotasks

Scale AI

Top Subject Matter

LLM evaluation in Japanese

Map search evaluation

AI generative response evaluation

Top Data Types

Image

Text

Top Task Types

Evaluation Rating

Prompt Response Writing SFT

Freelancer Overview

I participated several MK project Human evaluator in a Rainforcement Learning from Human Feedback (RLHF) Project: reviewed AI model response to various prompts. Scored or ranked according to quality, helpfulness, safety, and alignment with human expectations. Supervised/ Self-supervised ML Data work; Contributed to dataset creation and evaluation by: Assessing may serch results for accuracy and relevance. Selecting images that best matched user queries. Recording voice samples from scripts to support speech recognition and synthesis models.

Labeling Experience

RLHF experience

Scale AIAudioRLHFEvaluation Rating

Evalutaed and ranked AI-generated responses in verious areas, including helpfulness, correctness, and safety. The project was including code-mixing, language learning, occationally requires the use of extension, or image input.

Evalutaed and ranked AI-generated responses in verious areas, including helpfulness, correctness, and safety. The project was including code-mixing, language learning, occationally requires the use of extension, or image input.

2024 - 2025

Education

Work History