Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Factual Reasoning on MMLU Global Facts
Loading...
3
Baseline Wins
Domain-Grounded Tiered Retrieval
2.85
2.925
3
3.075
Mar 18, 2026
Baseline Wins
Proposed Wins
Tie Count
Win Rate
Hallucination Rate
Groundedness
Updated 1mo ago
Evaluation Results
Method
Method
Links
Baseline Wins
Proposed Wins
Tie Count
Win Rate
Hallucination Rate
Groundedness
Domain-Grounded Tiered Retrieval
N=50
2026.03
3
39
8
78
33.1
66.9
Baseline (Llama 3.1 8B)
N=50
2026.03
3
-
-
-
-
-
Feedback
Search any
task
Search any
task