Abstract

The present disclosure discloses an improved system and method for GenAI/LLM enabled payload generator for feature enrichment, selection and optimization. The present disclosure discloses a pipeline to apply fine-tuned large language model (LLM) to help feature simulators to test system performance and robustness. These templates may be incorporated into a multithreaded emulator to produce payloads that encompass a wide range of feature distributions, including high-volume and outlier cases. The Large Language Model (LLM) may be fine-tuned using a reinforcement learning feedback loop to detect potential issues in the pipeline. If a discrepancy arises between the business description and the feedback, a GenAI-enabled adversarial attack may be employed to test corner cases, ensuring robustness and alignment with business objectives.

Creative Commons License

Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.

Share

COinS