Article

Impact of Non-Standard Unicode Characters on Security and Comprehension in Large Language Models

2024-12-31

Abstract excerpt

<title>Abstract</title> <p>In this report, we present a comparative analysis of the performance of 15 large language models (LLMs). Each model underwent a standardized test comprising 38 queries obfuscated using alphanumeric symbols of non-standard Unicode characters. The responses of each model were analyzed using three key metrics: jailbreaks (prompt injections that cause a large language model to follow instru...

Topics

Open a Topic to create a Post that cites this publication.

COVID-19 diagnosis using AI

Identifiers and source

Literature Corpus work: 429c1259-8107-5a56-997e-1cafe2fe9c11
DOI: 10.21203/rs.3.rs-5723808/v1

Open publication

Related research

Semantic proximity does not establish scientific evidence.

Click a neighbor to travelStep 1 · 12 closest

Impact of Non-Standard Unicode Characters on Security and Comprehension in Large Language ModelsDOI 10.21203/rs.3.rs-5723808/v1Read

Knowledge Accuracy and Reducing Hallucinations in LLMs via Dynamic Domain Knowledge InjectionSemantic proximity Read

Drug or Pokémon? Large language model performance in identification of fabricated medicationsSemantic proximity Read

Evaluating the Quality of Large Language Model-Generated Cybersecurity Advice in GRC SettingsSemantic proximity Read

Benchmarking of Commercial Large Language Models: ChatGPT, Mistral, and LlamaSemantic proximity Read

DistilBERT: A Novel Approach to Detect Text Generated by Large Language Models (LLM)Semantic proximity Read

Applications of LLMs for Generating Cyber Security Exercise ScenariosSemantic proximity Read

A Survey on Privacy-Preserving Large LanguageModels (LLMs): Architectures, Techniques, and Use-CasesSemantic proximity Read

A Categorical Archive of ChatGPT FailuresSemantic proximity Read

An Evaluation of the Safety of ChatGPT with Malicious Prompt InjectionSemantic proximity Read

Synthetic Participants Generated by Large Language Models: A Systematic Literature ReviewSemantic proximity Read

Cognitive Integrity Security: Securing Human-AI Interaction in Large Language Model SystemsSemantic proximity Read

HyperLLM: The Next Generation of Large Language Models with Multimodal CapabilitiesSemantic proximity Read