Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Tool-use Reasoning on ShopTrajQA 64k context

60.1Accuracy

Qwen3-4B-SFT-RLVR

7.26820.98434.748.416Jun 6, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.06
60.1
2026.06
57.2
2026.06
51
2026.06
47.6
2026.06
41
2026.06
35.5
2026.06
27.7
2026.06
24.7
2026.06
22.4
2026.06
21.4
2026.06
18.8
2026.06
14.7
2026.06
9.3