Evaluating Language Model Safety Across Long Adversarial Conversations
cs.CL
Submitted: 2026-09-29
Updated: 2026-09-29
Code: https://github.com/meta-llama/PurpleLlama
Terminology
Sources
- Constitutional AI: Harmlessness from AI Feedback
- A Representation Engineering Perspective on the Effectiveness of Multi-Turn Jailbreaks
- Jailbreaking Black Box Large Language Models in Twenty Queries
- Llama Guard: LLM-based Input-Output Safeguard for Human-AI Conversations
- Capabilities of GPT-4 on Medical Challenge Problems
- LLMs know their vulnerabilities: Uncover Safety Gaps through Natural Distribution Shifts
- Scalable and Transferable Black-Box Jailbreaks for Language Models via Persona Modulation
- SimpleSafetyTests: a Test Suite for Identifying Critical Safety Risks in Large Language Models
- GPTFUZZER: Red Teaming Large Language Models with Auto-Generated Jailbreak Prompts
- ShieldGemma: Generative AI Content Moderation Based on Gemma
- Universal and Transferable Adversarial Attacks on Aligned Language Models
Related papers
- Exploring Solution Divergence and Its Effect on Large Language Model Problem Solving
- Ishigaki-IDS-Bench: A Benchmark for Generating Information Delivery Specification from BIM Information Requirements
- Subliminal Steering: Stronger Encoding of Hidden Signals
- MedStruct-S: A Benchmark for Key Discovery, Key-Conditioned QA and Semi-Structured Extraction from OCR Clinical Reports
- The End of Transformers? On Challenging Attention and the Rise of Sub-Quadratic Architectures
- Untangling the Mechanisms of Misleading Context in Medical Question Answering