Our new X account is live! Follow @wizwand_team for updates

Knowledge-Intensive Reasoning on Bamboogle (F1 score)

73.8F1

Llama3.1-8B + ARPO

Updated 2d ago

Evaluation Results

Method	Links
Llama3.1-8B + ARPO 2025.12		73.8
Qwen2.5-7B + ARPO 2025.12		71.5
Llama3.1-8B + Reinforce ++ 2025.12		69.1
Qwen2.5-7B + GRPO 2025.12		68.4
Llama3.1-8B + GRPO 2025.12		68.2
Llama3.1-8B + DAPO 2025.12		67.3
Qwen2.5-3B + ARPO 2025.12		66.8
Qwen2.5-3B + Reinforce ++ 2025.12		65.7
Qwen2.5-7B + DAPO 2025.12		65.5
Qwen2.5-3B + GRPO 2025.12		65.2
Qwen2.5-7B + Reinforce ++ 2025.12		64.9
Qwen2.5-3B + DAPO 2025.12		64.8
Llama3.1-8B + TIR Prompting 2025.12		58.4
Llama3.1-8B 2025.12		40
Qwen2.5-7B 2025.12		24
Qwen2.5-7B + TIR Prompting 2025.12		23.6
Qwen2.5-3B + TIR Prompting 2025.12		16.4
Qwen2.5-3B 2025.12		11.7