HazardWeaver: Scientific Route Selection for Hazard Analysis Agents

HazardWeaver is an AI agent designed to select and execute appropriate scientific methods for hazard analysis by leveraging a Hazard Knowledge Compiler and a Hazard Capability Graph. It dynamically revises its route choices as new evidence and execution results emerge, ensuring that selected methods are both applicable and executable with available data and tools. The system is evaluated on the Hazard Weaver Benchmark, which includes 141 instances across multiple hazard domains, and demonstrates superior performance over existing agent systems, especially when multiple scientific routes are viable.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv AI.

• llms
• agents
• benchmarks