Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Conversational Agent Evaluation on tau2 Airline (Success Rate)
Loading...
78
Success Rate
Pass@N (oracle)
26
39.5
53
66.5
Jun 24, 2026
Success Rate
Updated 1mo ago
Evaluation Results
Method
Method
Links
Success Rate
Pass@N (oracle)
Training=false, Backbo...
2026.06
78
DeepConf B10
Training=false, Backbo...
2026.06
72
Progress Advantage
Training=false, Backbo...
2026.06
72
Mean-of-N
Training=false, Backbo...
2026.06
64.8
WildReward-8B
Training=true, Backbon...
2026.06
64
ThinkPRM-7B
Training=true, Backbon...
2026.06
64
Self-Certainty
Training=false, Backbo...
2026.06
64
DeepConf Tail
Training=false, Backbo...
2026.06
62
Greedy Decoding
Training=false, Backbo...
2026.06
60
Pass@N (oracle)
Training=false, Backbo...
2026.06
58
ThinkPRM-14B
Training=true, Backbon...
2026.06
58
Progress Advantage
Training=false, Backbo...
2026.06
48
DeepConf Tail
Training=false, Backbo...
2026.06
36
Mean-of-N
Training=false, Backbo...
2026.06
34.5
Greedy Decoding
Training=false, Backbo...
2026.06
34
Self-Certainty
Training=false, Backbo...
2026.06
34
ThinkPRM-7B
Training=true, Backbon...
2026.06
30
WildReward-8B
Training=true, Backbon...
2026.06
28
ThinkPRM-14B
Training=true, Backbon...
2026.06
28
DeepConf B10
Training=false, Backbo...
2026.06
28
Feedback
Search any
task
Search any
task