HyperAI超神经

Text Summarization On Samsum Corpus

评估指标

ROUGE-1

ROUGE-L

评测结果

各个模型在此基准测试上的表现结果

			Paper Title
OmniVec2	59.1	63.7	OmniVec2 - A Novel Transformer based Network for Large Scale Multimodal and Multitask Learning
MoCa	55.13	50.88	Momentum Calibration for Text Generation
SICK	53.73	49.5	Mind the Gap! Injecting Commonsense Knowledge for Abstractive Dialogue Summarization
HAT-CNNDM RL	-	48.84	Hierarchical Learning for Generation with Long Source Sequences
InstructDS	55.3	46.7	Instructive Dialogue Summarization with Query Aggregations
PEGASUS 2B + SliC	54.37	45.89	Calibrating Sequence likelihood Improves Conditional Language Generation
ConDigSum	54.3	45.2	Topic-Aware Contrastive Learning for Abstractive Dialogue Summarization
BART-Large	52.6	42.1	Topic-Aware Contrastive Learning for Abstractive Dialogue Summarization
CriSPO 3-shot	47.2	38.2	CriSPO: Multi-Aspect Critique-Suggestion-guided Automatic Prompt Optimization for Text Generation
Mistral 7B + SigExt	44.1	33.9	Salient Information Prompting to Steer Content in Prompt-based Abstractive Summarization
GPT-4o (HPT)	-	30	Hierarchical Prompting Taxonomy: A Universal Evaluation Framework for Large Language Models
HAT-CNNDM	53.01	-	Hierarchical Learning for Generation with Long Source Sequences

0 of 12 row(s) selected.