Our new X account is live! Follow @wizwand_team for updates
Home
/
Benchmarks
Detoxification on SafeNLP (test)
Loading...
84
Similarity
Llama-2-7b
19.52
36.26
53
69.74
Jan 16, 2024
Similarity
Toxicity (%)
Updated 4d ago
Evaluation Results
Method
Method
Links
Similarity
Toxicity (%)
Llama-2-7b
Evaluation Mode=White-box
2024.01
84
76.9
Falcon-7b
Evaluation Mode=White-box
2024.01
66
58.9
GPT-3.5-Turbo
Evaluation Mode=White-box
2024.01
53
3.36
Mistral-7b
Evaluation Mode=White-box
2024.01
48
33.1
GPT-2-XL
Evaluation Mode=White-box
2024.01
46
28.18
Falcon-7b + CP
Evaluation Mode=White-...
2024.01
46
36.6
Mistral-7b + CP
Evaluation Mode=White-...
2024.01
40
4.3
GPT-2
Evaluation Mode=White-box
2024.01
36
28.94
CHRT
Evaluation Mode=White-...
2024.01
34
25.7
Distill-GPT-2
Evaluation Mode=White-box
2024.01
24
30.4
Model Arithmetic
Evaluation Mode=White-...
2024.01
24
12.2
Llama-2-7b + CP
Evaluation Mode=White-...
2024.01
24
11.4
CHRT
Evaluation Mode=White-...
2024.01
22
13.6
Feedback
Search any
task
Search any
task