Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Interactive Agent Task Completion on SocialMaze
Loading...
82.7
Average Reward
OpenSkill
65.228
69.764
74.3
78.836
Jun 4, 2026
Average Reward
Updated 1mo ago
Evaluation Results
Method
Method
Links
Average Reward
OpenSkill
Target Agent=Opus 4.6
2026.06
82.7
No Skill
Target Agent=Opus 4.6
2026.06
81.6
Skill Creator
Target Agent=Opus 4.6
2026.06
81
Self-Gen
Target Agent=Opus 4.6
2026.06
80.4
Memento
Target Agent=Opus 4.6
2026.06
80.3
CoT
Target Agent=Opus 4.6
2026.06
79.4
AutoSkill
Target Agent=Opus 4.6
2026.06
77.2
OpenSkill
Target Agent=GPT 5.2
2026.06
70.7
Skill Creator
Target Agent=GPT 5.2
2026.06
69.8
Self-Gen
Target Agent=GPT 5.2
2026.06
67.6
Memento
Target Agent=GPT 5.2
2026.06
67.4
CoT
Target Agent=GPT 5.2
2026.06
66.7
No Skill
Target Agent=GPT 5.2
2026.06
66.6
AutoSkill
Target Agent=GPT 5.2
2026.06
65.9
Feedback
Search any
task
Search any
task