REVA screens new functions for vulnerabilities in about one second
Original title: Reva: Real-Time Vulnerability Detection through Contextual Probing of Large Language Models
On PrimeVul, REVA with GLM-5.3-Flash matches VulTriage accuracy while running over 250 times faster.
Researchers present REVA, a detector that packs a module threat model, lightweight static-analysis findings and call-graph information into one LLM forward pass, then reads internal representations with a trained probe. On PrimeVul's held-out test set, REVA on GLM-5.3-Flash reaches 60.0% accuracy in about one second per function, matching VulTriage at 59.2% on the same model while costing less than a tenth as much. Across three open LLMs, the authors report REVA is at least as accurate as VulTriage.
Unconfirmed: Accuracy figures are the authors' reported results, not an independent evaluation.
- The probe is 14% to 17% more accurate than the LLM's own direct verdict.
- REVA is about 10% more accurate than VulTriage on two of three open LLMs.
- Authors say internal representations encode much of what agentic detectors gather by reasoning.
- AI models
- GLM-5.3-Flash
From the original article
1 min read at arxiv.orgA vulnerability costs more to fix the later it is discovered. As coding agents generate code faster than developers can review it, software needs a first line of defense screening each new function in real time.
By Amine Lbath · arxiv.org
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.