This is a binary classifier based on DeBERTa-v3-base for predicting if given a conversation, retrieval should be performed. It's trained on the ProCIS dataset.
@article{Anonymous_2024_ProCIS,
title = {ProCIS: A Benchmark for Proactive Retrieval in Conversations},
author = {},
journal = {Proceedings of the 47th International ACM SIGIR Conference on Research and Development in Information Retrieval},
year = {2024}
}