Fair Abstractive Summarization: Measuring and Improving Representation of Diverse Perspectives
Generating abstractive summaries that comprehensively cover diverse perspectives without underrepresenting certain groups.
COSMIC: Mutual Information for Task-Agnostic Summarization Evaluation
Die gegenseitige Information zwischen Quelltexten und Zusammenfassungen dient als universelles und aufgabenagnostisches Maß dafür, wie gut ein Zusammenfasser nützliche Informationen für Entscheidungen bewahrt.
COSMIC: Task-Agnostic Summarization Evaluation with Mutual Information
Mutual information between source texts and summaries is a task-agnostic measure of summarizer effectiveness.
Relation Triple-based Interpretable Summarization Framework: RTSUM
Leveraging relation triples for interpretable summarization.
Embrace Divergence for Richer Insights: Multi-document Summarization Benchmark and Case Study
LLMs struggle to cover diverse information effectively, highlighting the challenges in multi-document summarization.
Exploring Book-Length Summarization with BOOOOKSCORE
Automated metric BOOOOKSCORE evaluates book-length summarization coherence, revealing insights on LLM performance.
Information-Theoretic Distillation for Reference-less Summarization
Small-scale models can achieve competitive summarization results without relying on large language models or human-written references.
COSMIC: Mutual Information for Task-Agnostic Summarization Evaluation
The author proposes COSMIC as a task-oriented evaluation metric based on mutual information between source texts and summaries. It correlates well with human judgment-based metrics and predicts downstream task performance effectively.
© 2024 by Linnk AI