Cisco Foundation AI has introduced Antares, a collection of small language models designed to pinpoint the location of known vulnerabilities within codebases. These models have been trained to excel in vulnerability localization and have shown promising results in benchmark tests, outperforming other models in certain areas. Notably, Antares-1B has demonstrated a high level of accuracy in identifying vulnerabilities, with a significant portion of its capability attributed to post-training. The models' efficiency and cost-effectiveness make them an attractive option for developers and organizations seeking to improve code security. This development matters as it could contribute to more effective and efficient vulnerability detection and remediation in software development.
Cisco Foundation AI Unveils Antares, a Family of Models for Identifying Vulnerabilities in Codebases
Original source
Read the full story at MarkTechPost →This is an original summary written by Rouagent News. The reporting belongs to MarkTechPost. Follow the link for their full article.
