Multiple Ground Truths

The 'Multiple Ground Truths' dataset contains a collection of validated diverse responses to reduce noise and capture semantic diversity, and it is used to evaluate the effectiveness of adaptation methods for Large Language Models in technical service domains.

Papers using Multiple Ground Truths (1)

Lightweight Adaptation for LLM-based Technical Service Agent: Latent Logic Augmentation and Robust Noise Reduction2026