Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
PII Mitigation and Language Modeling on Web crawl prompt dataset (test)
Loading...
1,847.56
Avg PPL
Dedup
1,832.8744
1,932.0022
2,031.13
2,130.2578
Dec 2, 2025
Avg PPL
AURC
Updated 1mo ago
Evaluation Results
Method
Method
Links
Avg PPL
AURC
Dedup
Technique=Deduplication
2025.12
1,847.56
82.09
Baseline
Technique=Baseline
2025.12
2,168.39
-
RMFT
Technique=RMFT
2025.12
2,214.7
76.84
Feedback
Search any
task
Search any
task