notesum.ai
Published at November 18METEOR: Evolutionary Journey of Large Language Models from Guidance to Self-Growth
cs.AI
cs.CL
Released Date: November 18, 2024
Authors: Jiawei Li1, Chong Feng1, Yang Gao1
Aff.: 1Beijing Institute of Technology, Beijing, China
| Model | Accuracy | Completeness | Relevance | Coherence | Reliability | GPT4 Score | |
|---|---|---|---|---|---|---|---|
| LLaMA3 | w/o. | 21.3% | 27.8% | 25.6% | 19.9% | 17.8% | 5.02 |
| Meteor | 78.7% | 72.2% | 74.4% | 80.1% | 82.2% | 9.17 | |
| Qwen2 | w/o. | 31.6% | 36.5% | 39.7% | 36.5% | 32.9% | 6.88 |
| Meteor | 68.4% | 63.5% | 60.3% | 63.5% | c% | 9.28 | |