Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Preference Prediction on UltraFeedback 500 held-out users (test)
Loading...
70.53
Test Accuracy
RFM(32)
51.6436
56.5468
61.45
66.3532
Mar 21, 2025
Test Accuracy
Updated 1mo ago
Evaluation Results
Method
Method
Links
Test Accuracy
RFM(32)
n_hat=10
2025.03
70.53
Linear baseline
n_hat=10
2025.03
62.85
Baseline
n_hat=10
2025.03
62.58
Non-linear phi (3 layers)
n_hat=100
2025.03
57.06
Non-linear phi (5 layers)
n_hat=100
2025.03
56.2
Non-linear phi (3 layers)
n_hat=10
2025.03
55.43
Non-linear phi (5 layers)
n_hat=10
2025.03
52.37
Feedback
Search any
task
Search any
task