Show HN: Prompt injection detector beats ProtectAI by 19% accuracy, 8.9x smaller(huggingface.co)

3 pointsby Karan047vor 3 Monaten2 comments

Show HN: Prompt injection detector beats ProtectAI by 19% accuracy, 8.9x smaller

https://huggingface.co/hlyn/prompt-injection-judge-deberta-70m

ProtectAI's deberta-v3 is the most downloaded prompt injection classifier on HuggingFace. We benchmarked directly against it on Qualifire's independent dataset (excluded from our training).

* 91.68% vs 72.28% accuracy

* 95.84% vs 65.33% precision — they block 1 in 3 legitimate users

* 83MB vs 739MB ONNX size

* 101ms vs 646ms on CPU

3 comments

prabhkesarvor 3 Monaten

Prompt injection will never be solved. Instruction following is the product.

Karan047vor 3 Monaten

Nothing in cybersecurity is truly solved—only made harder, Prompt injection is no different.

JoshBlythevor 3 Monaten