List your company

Multi-hop Question Answering on 2WikiMultiHopQA online Google Search API (test val)

63.5Exact MatchESearch-base
34.3841.9449.557.06Oct 1, 2025

Evaluation results

MethodLinks
ESearch-baseModel Scale=7B2025.1063.573
ESearch-baseModel Scale=3B2025.1058.168.4
StepSearch-baseModel Scale=7B2025.1056.163.8
ESearch-instructModel Scale=7B2025.105565.4
Research-instructModel Scale=7B2025.1053.962.8
Search-R1-baseModel Scale=7B2025.1053.360.7
ZeroSearch-baseModel Scale=7B2025.1052.559.3
ESearch-instructModel Scale=3B2025.1052.164.4
R-Search-instruct-PPOModel Scale=3B2025.1049.655.8
StepSearch-instructModel Scale=7B2025.1048.557
Search-R1-instructModel Scale=7B2025.1047.556.1
R-Search-instruct-GRPOModel Scale=7B2025.104753.5
R-Search-instruct-PPOModel Scale=7B2025.1046.252.7
R-Search-instruct-GRPOModel Scale=3B2025.1045.751.9
Research-baseModel Scale=7B2025.1045.753.4
StepSearch-baseModel Scale=3B2025.1043.454.2
ZeroSearch-instructModel Scale=7B2025.1043.153.4
StepSearch-instructModel Scale=3B2025.1040.249.6
ZeroSearch-baseModel Scale=3B2025.1039.247.3
Search-R1-baseModel Scale=3B2025.1038.144.5
Search-R1-instructModel Scale=3B2025.1037.145.2
SSRL-instructModel Scale=3B2025.1036.546.1
SSRL-instructModel Scale=7B2025.1035.844.2
ZeroSearch-instructModel Scale=3B2025.1035.544.1