Text Fluency on CHED and CounterFact
16.6Average ScoreFT-M
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| FT-MBase Model=Llama3-8B-Instruct2025.05 | 16.6 | 128.9 | |
| PMETBase Model=Llama3-8B-Instruct2025.05 | 13.4 | 16.5 | |
| BaseBase Model=Llama3-8B-Instruct2025.05 | 13.3 | 11.1 | |
| CoRE-p+rBase Model=Llama3-8B-Instruct2025.05 | 13.2 | 13.3 | |
| MEMITBase Model=Llama3-8B-Instruct2025.05 | 13 | 13.1 | |
| CoRE-pBase Model=Llama3-8B-Instruct2025.05 | 13 | 14 | |
| JEEPBase Model=Llama3-8B-Instruct2025.05 | 12.8 | 21.8 | |
| PMETBase Model=Mistral-7B-Instruct2025.05 | 7.5 | 6.3 | |
| EMMETBase Model=Llama3-8B-Instruct2025.05 | 7.4 | 29.3 | |
| FT-MBase Model=Mistral-7B-Instruct2025.05 | 7.3 | 8.5 | |
| JEEPBase Model=Mistral-7B-Instruct2025.05 | 6.8 | 5.9 | |
| BaseBase Model=Mistral-7B-Instruct2025.05 | 6.7 | 6.3 | |
| CoRE-p+rBase Model=Mistral-7B-Instruct2025.05 | 6.6 | 6.2 | |
| EMMETBase Model=Mistral-7B-Instruct2025.05 | 6.4 | 6.4 | |
| CoRE-pBase Model=Mistral-7B-Instruct2025.05 | 6.4 | 6.1 | |
| MEMITBase Model=Mistral-7B-Instruct2025.05 | 5.8 | 6.1 |