Dolnapas Chonvasin

Key Skills

Software

Remotasks

Scale AI

Top Subject Matter

LLM evaluation in Thai/English

Prompt and Response Writing (Localization)

Voice-to-text transcription

Top Data Types

Document

Text

Top Label Types

Evaluation Rating

Prompt Response Writing SFT

Question Answering

Text Generation

Translation Localization

Freelancer Overview

- Verify information from credible web sources and evaluate AI-generated responses. - Offer detailed justifications for AI responses, ensuring accuracy and relevance through clear and comprehensive English. - Assist in voice-to-text transcription and data annotation.

Labeling Experience

Factuality and Creative Rewrite

Scale AITextPrompt Response Writing SFT

Create prompts that request the information in different styles, tones, or any criteria specified in the given guidelines and rubric, and then the contributor writes a response that meets the requirement of the prompt (ensure that all the information in the responses is accurate and up-to-date). An AI-generated response will be created for the prompt, side-by-side with a contributor’s responses. Select the best response and write a justification on why one response is better than the other.

Create prompts that request the information in different styles, tones, or any criteria specified in the given guidelines and rubric, and then the contributor writes a response that meets the requirement of the prompt (ensure that all the information in the responses is accurate and up-to-date). An AI-generated response will be created for the prompt, side-by-side with a contributor’s responses. Select the best response and write a justification on why one response is better than the other.

2024

AI Responses Factulity Evaluation

Scale AITextEvaluation Rating

Assess the prompt's suitability for evaluation (ratable) and proceed by answering two initial questions regarding two responses. If each response contains factual information, the evaluator then performs research to assess whether the given claims are correct. In the end, an evaluator compares two responses, determines the superior one, and provides a rationale for their choice.

Assess the prompt's suitability for evaluation (ratable) and proceed by answering two initial questions regarding two responses. If each response contains factual information, the evaluator then performs research to assess whether the given claims are correct. In the end, an evaluator compares two responses, determines the superior one, and provides a rationale for their choice.

2024 - 2024

Quality Evaluation of 2 AI Voice Responses

Scale AITextEvaluation Rating

- Evaluate the quality of two distinct AI-generated voice responses (GenAI chatbots) by comparing them. Begin by reading the user prompt in text format, followed by listening to both responses via a text-to-speech system. The objective of this project is to evaluate the quality of these responses and determine which one is superior based on six dimensions: overall quality, ease of hearing, verbosity, instruction following, truthfulness, and harmlessness. In the end, provides a rational justification for why one response is superior. - When deciding on the superior response, contemplate how effectively each GenAI formulates a response suited for voice or audio interactions, considering factors such as duration for listening and conversational tone.

- Evaluate the quality of two distinct AI-generated voice responses (GenAI chatbots) by comparing them. Begin by reading the user prompt in text format, followed by listening to both responses via a text-to-speech system. The objective of this project is to evaluate the quality of these responses and determine which one is superior based on six dimensions: overall quality, ease of hearing, verbosity, instruction following, truthfulness, and harmlessness. In the end, provides a rational justification for why one response is superior. - When deciding on the superior response, contemplate how effectively each GenAI formulates a response suited for voice or audio interactions, considering factors such as duration for listening and conversational tone.

2024 - 2024

Quality Evaluation of 2 AI Responses

Scale AITextEvaluation Rating

The project aims to evaluate and compare two AI responses to ascertain the accuracy, helpfulness, and engagement of each. This evaluation encompasses various quality criteria to determine the superior response, and then provide an overall justification.

The project aims to evaluate and compare two AI responses to ascertain the accuracy, helpfulness, and engagement of each. This evaluation encompasses various quality criteria to determine the superior response, and then provide an overall justification.

2024 - 2024

Image Generation Prompt Classification

Scale AITextClassification

Read the prompt and the provided category, then assess whether the prompt aligns correctly with the category as per the instructions.

Read the prompt and the provided category, then assess whether the prompt aligns correctly with the category as per the instructions.

2024 - 2024

Education

Work History