SOTAVerified

Red Teaming

Papers

Recently AddedMost HypedMost ActiveNeeds VerificationMost Verified

Showing 176–200 of 251 papers

TitleStatusHype
Against The Achilles' Heel: A Survey on Red Teaming for Generative ModelsCode2
Aurora-M: Open Source Continual Pre-training for Multilingual Language and Code—0
IterAlign: Iterative Constitutional Alignment of Large Language Models—0
HRLAIF: Improvements in Helpfulness and Harmlessness in Open-domain Reinforcement Learning From AI Feedback—0
Distract Large Language Models for Automatic Jailbreak AttackCode0
Red Teaming Models for Hyperspectral Image Analysis Using Explainable AI—0
Defending Against Unforeseen Failure Modes with Latent Adversarial TrainingCode1
Aligners: Decoupling LLMs and AlignmentCode0
A Safe Harbor for AI Evaluation and Red Teaming—0
Curiosity-driven Red-teaming for Large Language ModelsCode2
AttackGNN: Red-Teaming GNNs in Hardware Security Using Reinforcement Learning—0
Adversarial Nibbler: An Open Red-Teaming Method for Identifying Diverse Harms in Text-to-Image GenerationCode1
Agent Smith: A Single Image Can Jailbreak One Million Multimodal LLM Agents Exponentially FastCode2
HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust RefusalCode4
Investigating Bias Representations in Llama 2 Chat via Activation Steering—0
Gradient-Based Language Model Red TeamingCode0
Towards Red Teaming in Multimodal and Multilingual Translation—0
Red-Teaming for Generative AI: Silver Bullet or Security Theater?—0
Digital cloning of online social networks for language-sensitive agent-based modeling of misinformation spread—0
Red Teaming Visual Language Models—0
Sowing the Wind, Reaping the Whirlwind: The Impact of Editing Language ModelsCode0
Red Teaming for Large Language Models At Scale: Tackling Hallucinations on Mathematics TasksCode0
Causality Analysis for Evaluating the Security of Large Language ModelsCode1
AI Control: Improving Safety Despite Intentional SubversionCode1
Control Risk for Potential Misuse of Artificial Intelligence in ScienceCode1
Show:102550
← PrevPage 8 of 11Next →

Benchmark Results

#ModelMetricClaimedVerifiedStatus
1SUDOAttack Success Rate41—Unverified