Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Preference Alignment on UltraFeedback (test)
Loading...
74.18
Accuracy
FedPDPO
68.9176
70.2838
71.65
73.0162
Mar 20, 2026
Accuracy
Updated 26d ago
Evaluation Results
Method
Method
Links
Accuracy
FedPDPO
Backbone=TinyLlama-1.1...
2026.03
74.18
FedPer(+DPO)
Backbone=TinyLlama-1.1...
2026.03
73.12
FedRep(+DPO)
Backbone=TinyLlama-1.1...
2026.03
72.67
Per-FedAvg(+DPO)
Backbone=TinyLlama-1.1...
2026.03
72.41
FedAMP(+DPO)
Backbone=TinyLlama-1.1...
2026.03
71.56
FedPer(+PPO)
Backbone=TinyLlama-1.1...
2026.03
71.38
FedAvg(+DPO)
Backbone=TinyLlama-1.1...
2026.03
71.28
FedRep(+PPO)
Backbone=TinyLlama-1.1...
2026.03
70.29
Per-FedAvg(+PPO)
Backbone=TinyLlama-1.1...
2026.03
70.15
FedAvg(+PPO)
Backbone=TinyLlama-1.1...
2026.03
69.54
FedAMP(+PPO)
Backbone=TinyLlama-1.1...
2026.03
69.12
Feedback
Search any
task
Search any
task