Learning Agentic Policy from Action Guidance
Open paper
Benchmark
Model
Rank
Results
agents-on-gaia-text-only-103-validation-tasks
ActGuide-RL (Qwen3-8B)
#7
Accuracy: 41.74
deep-research-agents-on-browsecomp-zh
ActGuide-RL (Qwen3-8B)
#11
Accuracy: 26.64
Rank counts only results with a code link.