Our new X account is live! Follow @wizwand_team for updates
Home
/
Benchmarks
Workflow Reconstruction on SCI (test)
Loading...
0.139
SFE
SFT
0.12828
0.20064
0.273
0.34536
Feb 5, 2026
SFE
Updated 4d ago
Evaluation Results
Method
Method
Links
SFE
SFT
Protocol=Direct input-...
2026.02
0.139
Claude Opus 4.5
Protocol=Multi-turn se...
2026.02
0.299
ReAct (Claude Opus 4.5)
Protocol=ReAct-style t...
2026.02
0.305
AgentXRay w/o Pruning
Ablation=Without Pruning
2026.02
0.312
AFlow
Protocol=MCTS-based wo...
2026.02
0.373
AgentXRay w/o Tools
Ablation=Without Tools
2026.02
0.378
AgentXRay (All Tools)
Config=Full primitive...
2026.02
0.395
AgentXRay (Selected)
Config=Selected primit...
2026.02
0.407
Feedback
Search any
task
Search any
task