VulTriage
An LLM-agent-based vulnerability detector that gathers context through multi-turn reasoning and tool calls to perform security analysis; it achieves high accuracy but incurs relatively high inference latency.
Timeline (last 90 days)
On PrimeVul, the agent-style detector VulTriage (also running on GLM-5.3-Flash) achieves 59.2% accuracy.
With precision on par with VulTriage, REVA is more than 250 times faster and costs less than one tenth as much.
In cross-model validation, REVA's accuracy is at least on par with VulTriage on three open-source LLMs, and about 10% higher on two of them.
All Facts (3)
On PrimeVul, the agent-style detector VulTriage (also running on GLM-5.3-Flash) achieves 59.2% accuracy.
50%UnverifiedWith precision on par with VulTriage, REVA is more than 250 times faster and costs less than one tenth as much.
50%UnverifiedIn cross-model validation, REVA's accuracy is at least on par with VulTriage on three open-source LLMs, and about 10% higher on two of them.
50%