Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Multiple-choice factual recall on POLYFACT High
Loading...
73.15
Accuracy
GRPO
55.6572
60.1986
64.74
69.2814
Jun 4, 2026
Accuracy
Updated 1mo ago
Evaluation Results
Method
Method
Links
Accuracy
GRPO
Backbone=Qwen-2.5-7B
2026.06
73.15
Baseline
Backbone=Qwen-2.5-7B
2026.06
66.69
GRPO
Backbone=OLMo-2-1124-7B
2026.06
64.21
CPT + GRPO
Backbone=Qwen-2.5-7B
2026.06
64.11
SFT
Backbone=Qwen-2.5-7B
2026.06
63.25
CPT
Backbone=Qwen-2.5-7B
2026.06
61.29
CPT + GRPO
Backbone=OLMo-2-1124-7B
2026.06
61.26
CPT + SFT
Backbone=OLMo-2-1124-7B
2026.06
59.2
CPT + SFT
Backbone=Qwen-2.5-7B
2026.06
58.1
Baseline
Backbone=OLMo-2-1124-7B
2026.06
57.93
CPT
Backbone=OLMo-2-1124-7B
2026.06
57.89
SFT
Backbone=OLMo-2-1124-7B
2026.06
56.33
Feedback
Search any
task
Search any
task
