PolicyGuide: From Guarding One Action to Guiding the Whole Workflow for Policy-Compliant LLM Agents
Open paper
Benchmark
Model
Rank
Results
agents-on-bench-airline
PolicyGuide (GPT-5.4)
#3
Accuracy: 77.5
agents-on-bench-retail
PolicyGuide (GPT-5.4)
#3
Accuracy: 80.9
agents-on-bench-telecom
PolicyGuide (GPT-5.4)
#5
Accuracy: 86.6
Rank counts only results with a code link.