REVIEW 7 cited by
Embedding-based classifiers can detect prompt injection attacks
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
read the original abstract
Large Language Models (LLMs) are seeing significant adoption in every type of organization due to their exceptional generative capabilities. However, LLMs are found to be vulnerable to various adversarial attacks, particularly prompt injection attacks, which trick them into producing harmful or inappropriate content. Adversaries execute such attacks by crafting malicious prompts to deceive the LLMs. In this paper, we propose a novel approach based on embedding-based Machine Learning (ML) classifiers to protect LLM-based applications against this severe threat. We leverage three commonly used embedding models to generate embeddings of malicious and benign prompts and utilize ML classifiers to predict whether an input prompt is malicious. Out of several traditional ML methods, we achieve the best performance with classifiers built using Random Forest and XGBoost. Our classifiers outperform state-of-the-art prompt injection classifiers available in open-source implementations, which use encoder-only neural networks.
Forward citations
Cited by 7 Pith papers
-
From Neural Intent to Cryptographic Authorization: Securing AI-Driven Enterprise Workflows
NCS gates every privileged AI-agent tool call behind an offline-signed, hash-chained instruction stream, cutting injected-parameter attacks to zero by construction.
-
Context Contamination in LLM Analysis of Network Security Logs: Poison with Passive Prompt Injection and Mitigation Evaluation
Malicious text hidden in log fields hijacks LLM-based security analysis up to 88.2% of the time, and layered defenses reduce but do not eliminate the risk.
-
Evaluating the Robustness of Retrieval-Augmented Generation to Adversarial Evidence in the Health Domain
Misleading health documents in RAG context sharply lower LLM accuracy, and heavily helpful-biased retrieval pools restore it.
-
A Critical Evaluation of Defenses against Prompt Injection Attacks
StruQ, SecAlign, Instruction Hierarchy, PromptGuard, and Attention Tracker are substantially less effective and utility-preserving than claimed when evaluated with diverse prompts and adaptive attacks.
-
CASE-Bench: Context-Aware SafEty Benchmark for Large Language Models
A new benchmark shows that human safety judgments about LLM responses shift strongly with context, and that current LLMs, especially commercial ones, often fail to match those judgments.
-
Zero-Shot Embedding Drift Detection: A Lightweight Defense Against Prompt Injections in LLMs
An embedding-drift prompt injection detector that is not zero-shot, requires a clean reference prompt at inference, and fits its threshold on the test set, so the reported >93% accuracy is not evidence of deployed per...
-
When Your Reviewer is an LLM: Biases, Divergence, and Prompt Injection Risks in Peer Review
GPT-5-mini gives weaker papers systematically higher scores than human reviewers, and hidden field-specific prompts in PDFs can force it to assign perfect scores or suppress weaknesses.
Discussion (0). Continue with ORCID to comment.