王瑞琦

Key Skills

Software

No software listed

Top Subject Matter

Empathy dialogue

model alignment

NLP decoding strategies

Top Data Types

Text

Top Task Types

RLHF

Freelancer Overview

NLP research papers (Inference-Time Pareto-Guided Alignment) — Person in charge. Brings 6+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include Internal and Proprietary Tooling. Education includes Bachelor of Science, Nanjing University (2022) and Master of Science, Kunming University of Science and Technology (2028). AI-training focus includes data types such as Text and labeling workflows including RLHF.

Labeling Experience

NLP research papers (Inference-Time Pareto-Guided Alignment) — Person in charge

TextRLHF

Worked as the person in charge for an NLP research project on empathy dialogue alignment. Focused on reducing distribution collapse in parameter-based alignment approaches such as RLHF-style methods. Implemented an inference-time strategy that generates multiple candidate responses and filters Pareto-optimal outputs using a multi-target discriminator. • Generated multiple candidate replies for empathy dialogue inference. • Applied Pareto-guided selection among candidates. • Used a multi-target discriminator for response filtering. • Aimed to improve alignment behavior in dialogue tasks.

2025 - 2025

Education

Work History