Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Deep search QA on WebWalkerQA
Loading...
23.01
Accuracy
ProCeedRL
18.4444
19.6297
20.815
22.0003
Apr 2, 2026
Accuracy
Updated 16d ago
Evaluation Results
Method
Method
Links
Accuracy
ProCeedRL
Backbone=Qwen3-8B
2026.04
23.01
Rewinding More
Backbone=Qwen3-8B, Typ...
2026.04
20.16
Qwen3-8B-v3-SFT
Backbone=Qwen3-8B
2026.04
19.85
RFT
Backbone=Qwen3-8B
2026.04
19.56
DAPO/Search-R1
Backbone=Qwen3-8B
2026.04
19.51
ReAct Prompting
Backbone=Qwen3-8B
2026.04
18.62
Feedback
Search any
task
Search any
task