No article text was recorded for this version.
role-play vulnerability
Version 2 of 2134 papers194 conceptsfrozen 2026-10-04
Papers in this version
- Blog: Fraud-R1 : A Multi-Round Benchmark for Assessing the Robustness of LLM Against Augmented Fraud and Phishing Inducements
- Blog: Toxicity in chatgpt: Analyzing persona-assigned language models
- Blog: Are Personalized Stochastic Parrots More Dangerous? Evaluating Persona Biases in Dialogue Systems
- Blog: Scalable and Transferable Black-Box Jailbreaks for Language Models via Persona Modulation
- Blog: Picky LLMs and Unreliable RMs: An Empirical Study on Safety Alignment after Instruction Tuning
- Blog: Beware of Your Po! Measuring and Mitigating AI Safety Risks in Role-Play Fine-Tuning of LLMs
- Blog: On the Decision-Making Abilities in Role-Playing using Large Language Models
- Blog: Fraud-R1 : A Multi-Round Benchmark for Assessing the Robustness of LLM Against Augmented Fraud and Phishing Inducements
- Blog: Red teaming ChatGPT via Jailbreaking: Bias, Robustness, Reliability and Toxicity
- Blog: FraudShield: Knowledge Graph Empowered Defense for LLMs against Fraud Attacks
- Blog: Exploring AI Ethics of ChatGPT: A Diagnostic Analysis
- Blog: Dr. Jekyll and Mr. Hyde: Two Faces of LLMs
- Blog: Evaluation of LLM Vulnerabilities to Being Misused for Personalized Disinformation Generation
- Blog: Disentangling Intent from Role: Adversarial Self-Play for Persona-Invariant Safety Alignment
- Blog: Exploiting Programmatic Behavior of LLMs: Dual-Use Through Standard Security Attacks
- Blog: BadRobot: Jailbreaking Embodied LLM Agents in the Physical World
- Blog: Fighting Fire with Fire: The Dual Role of LLMs in Crafting and Detecting Elusive Disinformation
- Blog: Scalable and Transferable Black-Box Jailbreaks for Language Models via Persona Modulation
- Blog: Who's asking? User personas and the mechanics of latent misalignment
- Blog: PIG: Privacy Jailbreak Attack on LLMs via Gradient-based Iterative In-Context Optimization
- Blog: Persona‑Conditioned Adversarial Prompting (PCAP): Multi‑Identity Red‑Teaming for Enhanced Adversarial Prompt Discovery
- Blog: No Two Devils Alike: Unveiling Distinct Mechanisms of Fine-tuning Attacks
- Blog: X-Teaming Evolutionary M2S: Automated Discovery of Multi-turn to Single-turn Jailbreak Templates
- Blog: Beware of Your Po! Measuring and Mitigating AI Safety Risks in Role-Play Fine-Tuning of LLMs
- Blog: Toxicity in ChatGPT: Analyzing Persona-assigned Language Models
- Blog: PsySafe: A Comprehensive Framework for Psychological-based Attack, Defense, and Evaluation of Multi-agent System Safety
- Blog: LLM Echo Chamber: personalized and automated disinformation
- Blog: Hey GPT, Can You be More Racist? Analysis from Crowdsourced Attempts to Elicit Biased Content from Generative AI
- Blog: Jailbreaking ChatGPT via Prompt Engineering: An Empirical Study
- Blog: Fine-tuning Aligned Language Models Compromises Safety, Even When Users Do Not Intend To!
- Blog: "Do Anything Now": Characterizing and Evaluating In-The-Wild Jailbreak Prompts on Large Language Models
- Blog: Summon a demon and bind it: A grounded theory of LLM red teaming
- Blog: Countering Autonomous Cyber Threats
- Blog: Mimicking User Data: On Mitigating Fine-Tuning Risks in Closed Large Language Models
- Blog: Smoke and Mirrors: Jailbreaking LLM-based Code Generation via Implicit Malicious Prompts
- Blog: Don't Listen To Me: Understanding and Exploring Jailbreak Prompts of Large Language Models
- Blog: Ethical Considerations and Policy Implications for Large Language Models: Guiding Responsible Development and Deployment
- Blog: AREG: Adversarial Resource Extraction Game for Evaluating Persuasion and Resistance in Large Language Models
- Blog: A Hitchhiker's Guide to Jailbreaking ChatGPT via Prompt Engineering
- Blog: Seeing Seeds Beyond Weeds: Green Teaming Generative AI for Beneficial Uses
- Blog: SoK: Prompt Hacking of Large Language Models
- Blog: Exposing AI Bias by Crowdsourcing: Democratizing Critique of Large Language Models
- Blog: The Compliance Paradox: Semantic-Instruction Decoupling in Automated Academic Code Evaluation
- Blog: How Well Do Models Follow Their Constitutions?
- Blog: ControlLM: Crafting Diverse Personalities for Language Models
- Blog: EMNLP: Educator-role Moral and Normative Large Language Models Profiling
- Blog: BadRobot: Jailbreaking LLM-based Embodied AI in the Physical World
- Blog: IMPersona: Evaluating Individual Level LLM Impersonation
- Blog: “What’s Up, Doc?”: Analyzing How Users Seek Health Information in Large-Scale Conversational AI Datasets
- Blog: P4: Plug-and-Play Discrete Prompting for Large Language Models Personalization
- Blog: Red-Teaming LLM Multi-Agent Systems via Communication Attacks
- Blog: MIRAGE: Multimodal Immersive Reasoning and Guided Exploration for Red-Team Jailbreak Attacks
- Blog: PersonaTeaming: Exploring How Introducing Personas Can Improve Automated AI Red-Teaming
- Blog: SoK: Large Language Model-Generated Textual Phishing Campaigns End-to-End Analysis of Generation, Characteristics, and Detection
- Blog: Consistency Training Along the Transformer Stack
- Blog: Voice Jailbreak Attacks Against GPT-4o
- Blog: PERSONA: Dynamic and Compositional Inference-Time Personality Control via Activation Vector Algebra
- Blog: SafetyQuizzer: Timely and Dynamic Evaluation on the Safety of LLMs
- Blog: PICLe: Eliciting Diverse Behaviors from Large Language Models with Persona In-Context Learning
- Blog: Eliciting Secret Knowledge from Language Models
Showing 60 of 134 papers.
History & citation
Current · Permanent link
Cite this direction
Latest version (a living page):
@misc{lacuna_role_play_vulnerability_1c2ea6cc,
title = {role-play vulnerability},
author = {{Lacuna}},
howpublished = {Lacuna research direction},
year = {2026},
url = {https://lacuna.tiptreesystems.com/direction/role-play-vulnerability/rdr_1c2ea6cc73be4793aa3123ec338682ed},
note = {Latest version of a living research direction; cite a version for a fixed text. Accessed 2026-11-20},
}This exact version (version 2, frozen text):
@misc{lacuna_role_play_vulnerability_1c2ea6cc_v2,
title = {role-play vulnerability},
author = {{Lacuna}},
howpublished = {Lacuna research direction},
year = {2026},
url = {https://lacuna.tiptreesystems.com/direction/role-play-vulnerability/rdr_1c2ea6cc73be4793aa3123ec338682ed/v/2},
note = {Version 2, frozen 2026-10-04. Accessed 2026-11-20},
}