Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Fact Generation (Targeted) on WD50K
Loading...
50.8
Accuracy
KREPE
9.616
20.308
31
41.692
May 22, 2026
Accuracy
Updated 8d ago
Evaluation Results
Method
Method
Links
Accuracy
KREPE
LLM=N/A
2026.05
50.8
Random Facts
LLM=Gemini 3 Pro
2026.05
48.2
Random Facts
LLM=GPT 5.2
2026.05
46.3
Neighbor Sets
LLM=Gemini 3 Pro
2026.05
44.5
Neighbor Sets
LLM=GPT 5.2
2026.05
39.7
Few-shot Facts
LLM=Gemini 3 Pro
2026.05
28.1
Iterative Prediction
LLM=N/A
2026.05
27
Re-ranking
LLM=Gemini 3 Pro
2026.05
25.6
Autoregressive
LLM=GPT 5.2
2026.05
25
Few-shot Facts
LLM=GPT 5.2
2026.05
22.2
Gibbs Sampling
LLM=N/A
2026.05
19.9
Autoregressive
LLM=Gemini 3 Pro
2026.05
19
Re-ranking
LLM=GPT 5.2
2026.05
11.2
Feedback
Search any
task
Search any
task