Neuralchemy

AI Security • Prompt Defense • LLM Safety

Building secure, reliable AI systems focused on prompt security, adversarial robustness, and practical safety tooling.

Featured Project — PromptShield

Prompt injection detection framework designed to classify benign vs malicious prompts across real-world and synthetic attack patterns.

Resources

• Dataset: neuralchemy/prompt-injection-dataset 6000+ prompt injection and benign samples collected from realistic attack scenarios.

• ML Models: neuralchemy/prompt-injection-detector Classical machine learning classifiers for prompt risk detection.

• DeBERTa Fine-Tuned Model: neuralchemy/prompt-injection-deberta Transformer-based prompt injection classifier.

• Live Demo Space: Prompt-injection-DeBERTa Interactive inference demo for prompt safety classification.

Mission

Advancing AI security through open datasets, practical model deployment, and adversarial safety research.

Focus Areas

• Prompt Injection Detection
• Adversarial Prompt Benchmarking
• LLM Safety Evaluation
• Secure AI Tooling

Links

• Website: https://www.neuralchemy.in
• GitHub: https://github.com/neuralchemy

Building safer AI systems through open security research. 🚀