Picture for Yixiao Song

Yixiao Song

VERISCORE: Evaluating the factuality of verifiable claims in long-form text generation

Add code
Jun 27, 2024
Viaarxiv icon

GEE! Grammar Error Explanation with Large Language Models

Add code
Nov 16, 2023
Viaarxiv icon

A Critical Evaluation of Evaluations for Long-form Question Answering

Add code
May 29, 2023
Figure 1 for A Critical Evaluation of Evaluations for Long-form Question Answering
Figure 2 for A Critical Evaluation of Evaluations for Long-form Question Answering
Figure 3 for A Critical Evaluation of Evaluations for Long-form Question Answering
Figure 4 for A Critical Evaluation of Evaluations for Long-form Question Answering
Viaarxiv icon

KNN-LM Does Not Improve Open-ended Text Generation

Add code
May 24, 2023
Figure 1 for KNN-LM Does Not Improve Open-ended Text Generation
Figure 2 for KNN-LM Does Not Improve Open-ended Text Generation
Figure 3 for KNN-LM Does Not Improve Open-ended Text Generation
Figure 4 for KNN-LM Does Not Improve Open-ended Text Generation
Viaarxiv icon

Paraphrasing evades detectors of AI-generated text, but retrieval is an effective defense

Add code
Mar 23, 2023
Viaarxiv icon

DEMETR: Diagnosing Evaluation Metrics for Translation

Add code
Oct 25, 2022
Figure 1 for DEMETR: Diagnosing Evaluation Metrics for Translation
Figure 2 for DEMETR: Diagnosing Evaluation Metrics for Translation
Figure 3 for DEMETR: Diagnosing Evaluation Metrics for Translation
Figure 4 for DEMETR: Diagnosing Evaluation Metrics for Translation
Viaarxiv icon