List your company
ResearchFormerly WizWand

Question Answering on PopQA out-of-domain

46.9Exact MatchSearch-R1-7B-Base-PPO w/ Format
3.7414.94526.1537.355Oct 15, 2025
Updated 5mo ago

Evaluation results

MethodLinks
Search-R1-7B-Base-PPO w/ FormatModel Scale=7B, Training Algorithm=PPO, Input Formatting=Included2025.1046.9
EM+Info-Think+Think-AnsFramework=VERITAS-R1, Reward Components=EM+Info-Think+Think-Ans2025.1046.6
EM+Think-AnsFramework=VERITAS-R1, Reward Components=EM+Think-Ans2025.1046.4
EM+Info-ThinkFramework=VERITAS-R1, Reward Components=EM+Info-Think2025.1046.3
Search-R1-7B-Base-PPOModel Scale=7B, Training Algorithm=PPO2025.1045.7
ReSearch-7B-InstructModel Scale=7B, Model Variant=Instruct2025.1045
ReSearch-7B-BaseModel Scale=7B, Model Variant=Base2025.1044.2
DeSA-7B-Instrct-GRPOModel Scale=7B, Training Algorithm=GRPO2025.1044
DeSA-3B-Instruct-GRPOModel Scale=3B, Training Algorithm=GRPO2025.1039.7
RAGReasoning Strategy=Retrieval-Augmented Generation2025.1039.2
Rejection SamplingReasoning Strategy=Rejection Sampling2025.1038
IRCoTReasoning Strategy=Iterative Retrieval Chain-of-Thought2025.1030.1
R1-baseModel Variant=Base2025.1020.2
R1-instructModel Variant=Instruct2025.1019.9
Direct InferenceReasoning Strategy=Direct Inference2025.1014
Search-o1Reasoning Strategy=Search-o12025.1013.1
SFTTraining Algorithm=Supervised Fine-Tuning2025.1012.1
CoTReasoning Strategy=Chain-of-Thought2025.105.4