Primer Research

This repository contains or refers to implementations, research code and data related to Primer publications and preprints.

Paper	Source/Data
How important is Recall for Measuring Retrieval Quality? 2025	recall_role, retrieval-response
Preserving Multilingual Quality While Tuning Query Encoder on English Only. 2025	arxiv-negatives
How to Discern Important Urgent News? 2024
Linear Cross-Lingual Mapping of Sentence Embeddings. 2024	wikinews (or here)
Neural Embeddings for Text. 2022	neural_embeddings
BabyBear: Cheap inference triage for expensive language models. 2022	babybear
Named Entity Linking with Entity Representation by Multiple Embeddings. 2022
Does Summary Evaluation Survive Translation to Other Languages? 2022	summary_translation
Play the Shannon Game with Language Models: A Human-Free Approach to Summary Evaluation. 2022	shannon, blanc
Consistency and Coherence from Points of Contextual Similarity. 2021	estime-soft, blanc
Namesakes: Ambiguously Named Entities from Wikipedia and News. 2021	namesakes
ESTIME: Estimation of Summary-to-Text Inconsistency by Mismatched Embeddings. 2021	estime, blanc
Towards Human-Free Automatic Quality Evaluation of German Summarization. 2021
Is Human Scoring the Best Criteria for Summary Evaluation? 2021
Primer AI's Systems for Acronym Identification and Disambiguation. 2021	sdu-data
Sensitivity of BLANC to human-scored qualities of text summaries. 2020	blanc, data
Fill in the BLANC: Human-free quality estimation of document summaries. 2020	blanc, data
Zero-shot topic generation. 2020
Headline Generation: Learning from Decomposable Document Titles. 2019