Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Search Relevance Evaluation on Taobao AI Search Tail Queries
Loading...
48.5
Hit Rate
LEAPS-HR
36.02
39.26
42.5
45.74
Jan 9, 2026
Hit Rate
GR
Error Rate
Updated 5mo ago
Evaluation Results
Method
Method
Links
Hit Rate
GR
Error Rate
LEAPS-HR
Base Model=Qwen3-14B,...
2026.01
48.5
61
51
LEAPS-ER
Base Model=Qwen3-14B,...
2026.01
47
64.7
54.3
LEAPS-GR
Base Model=Qwen3-14B,...
2026.01
45.3
66
52.3
SFT (Qwen3-14B)
Base Model=Qwen3-14B,...
2026.01
37.5
50.5
42.9
SFT (Tbstar-16A2B)
Base Model=Tbstar-16A2...
2026.01
37.4
50.3
43.1
Production Baseline
2026.01
36.5
48.8
41.8
Feedback
Search any
task
Search any
task