Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Tool-use Generalization on ACEBench Format and Content OOD v1 (test)
Loading...
38
Adherence (Ad.)
Vanilla
-1.52
8.74
19
29.26
Jun 24, 2026
Adherence (Ad.)
Success Rate (Sw.)
Step Fidelity (sf.)
Multi-Turn Score
Average Score
Updated 1mo ago
Evaluation Results
Method
Method
Links
Adherence (Ad.)
Success Rate (Sw.)
Step Fidelity (sf.)
Multi-Turn Score
Average Score
Vanilla
Backbone=Qwen2.5-1.5B-...
2026.06
38
20
46
0
26
OPS
Backbone=Qwen2.5-1.5B-...
2026.06
36
24
43
0
25.75
GRPO
Backbone=Qwen2.5-1.5B-...
2026.06
32
24
43
0
24.75
HBG
Backbone=Qwen2.5-1.5B-...
2026.06
28
20
42
0
22.5
PRS
Backbone=Qwen2.5-1.5B-...
2026.06
16
8
24
0
12
ETS
Backbone=Qwen2.5-1.5B-...
2026.06
10
0
25
0
8.75
SFT_BFCL
Backbone=Qwen2.5-1.5B-...
2026.06
0
0
0
0
0
SFT_BFCL + RL
Backbone=Qwen2.5-1.5B-...
2026.06
0
0
0
0
0
Feedback
Search any
task
Search any
task
