What Did the Agent Actually Do? Evidence-Grounded Oversight for Long-Horizon Agents
To support these judgments, we propose the Evidence-Grounded Behavior Graph (EBG), a training-free method that groups source-linked evidence into behaviors and organizes their relationships into a graph.
ProofPaper ↗
Key points
- As agents take on long-horizon tasks, users shift from making individual decisions to overseeing autonomous execution.
- Yet the volume of agent activity and the fragmentation of supporting evidence make it difficult to determine which decisions warrant user verification.
- We study monitors that identify consequential decisions and locate evidence to help users assess their implications.
- We introduce AgentMonBench, a software-engineering benchmark comprising three subsets that cover two complementary dimensions: alignment between requirements and behavior, and awareness of consequential autonomous decisions for verification.
Sources (1)
- [1]What Did the Agent Actually Do? Evidence-Grounded Oversight for Long-Horizon AgentsHugging Face Daily Papers · Oct 5, 12:00 AM
To support these judgments, we propose the Evidence-Grounded Behavior Graph (EBG), a training-free method that groups source-linked evidence into behaviors and organizes their relationships into a graph.
As agents take on long-horizon tasks, users shift from making individual decisions to overseeing autonomous execution.
Extractive summary: sentences quoted from the sources.