List your company

Multi-hop Question Answering on 2WikiMultiHopQA offline Wiki-18 (test val)

43.6Exact MatchESearch-base
23.11228.43133.7539.069Oct 1, 2025

Evaluation results

MethodLinks
ESearch-baseModel Scale=7B2025.1043.651.3
ESearch-baseModel Scale=3B2025.1042.849.9
ESearch-instructModel Scale=7B2025.1041.949.4
ESearch-instructModel Scale=3B2025.1041.550
StepSearch-baseModel Scale=7B2025.1038.545
StepSearch-instructModel Scale=7B2025.1036.643.1
ZeroSearch-instructModel Scale=7B2025.103642.2
Research-instructModel Scale=7B2025.1035.441.6
Search-R1-baseModel Scale=7B2025.103541.1
R-Search-instruct-GRPOModel Scale=7B2025.1034.640.1
StepSearch-baseModel Scale=3B2025.1033.939.5
SSRL-instructModel Scale=7B2025.1033.239.9
StepSearch-instructModel Scale=3B2025.103238.5
Search-R1-instructModel Scale=7B2025.1031.237.6
R-Search-instruct-GRPOModel Scale=3B2025.1030.735.1
ZeroSearch-baseModel Scale=7B2025.1029.735.6
Search-R1-instructModel Scale=3B2025.1029.335.2
SSRL-instructModel Scale=3B2025.102934.8
R-Search-instruct-PPOModel Scale=3B2025.1027.732.8
R-Search-instruct-PPOModel Scale=7B2025.1027.433.9
Research-baseModel Scale=7B2025.1026.431.3
ZeroSearch-baseModel Scale=3B2025.1025.331.1
Search-R1-baseModel Scale=3B2025.1024.829.6
ZeroSearch-instructModel Scale=3B2025.1023.928.8