FinRED creates an expert-validated benchmark and rubric for financial LLM safety that maps regulatory standards to specific threats and reduces critical false negatives in evaluation from 28 to 12.
Redagent: Red teaming large language models with context-aware au- tonomous language agent.arXiv preprint arXiv:2407.16667
7 Pith papers cite this work. Polarity classification is still indexing.
7
Pith papers citing it
citation-role summary
background 2
baseline 1
citation-polarity summary
representative citing papers
SentGuard achieves 90.5% detection of unsafe cases within two sentences at 7.41% false positive rate by operating at sentence boundaries during LLM streaming generation.
PCAP conditions adversarial searches on multiple attacker personas to discover more diverse and transferable jailbreaks, yielding richer safety fine-tuning datasets that boost model robustness on GPT-OSS 120B.
TurnGate identifies the critical turn in multi-turn dialogues where a response would complete hidden malicious intent, outperforming baselines on the new MTID dataset while keeping over-refusal low.
EvoSynth evolves code-based jailbreak algorithms via multi-agent self-correction, reaching 85.5% ASR on Claude-Sonnet-4.5 and 95.9% average across targets with greater diversity.
A multi-role red-teaming framework with attacker, target, and jury LLMs measures faithfulness in English and Arabic, finding false-premise prompts and length limits change unfaithfulness rates.
ADR is a three-component detection system for AI agents that combines telemetry sensors, red teaming, and two-tier detection, achieving 97.2% precision in a ten-month Uber deployment and outperforming baselines on the new ADR-Bench.
citing papers explorer
-
FinRED: An Expert-Guided Benchmark Generation and Evaluation Framework for Financial LLM Red-Teaming
FinRED creates an expert-validated benchmark and rubric for financial LLM safety that maps regulatory standards to specific threats and reduces critical false negatives in evaluation from 28 to 12.
-
SentGuard: Sentence-Level Streaming Guardrails for Large Language Models
SentGuard achieves 90.5% detection of unsafe cases within two sentences at 7.41% false positive rate by operating at sentence boundaries during LLM streaming generation.
-
Persona-Conditioned Adversarial Prompting: Multi-Identity Red-Teaming for Adversarial Discovery and Mitigation
PCAP conditions adversarial searches on multiple attacker personas to discover more diverse and transferable jailbreaks, yielding richer safety fine-tuning datasets that boost model robustness on GPT-OSS 120B.
-
One Turn Too Late: Response-Aware Defense Against Hidden Malicious Intent in Multi-Turn Dialogue
TurnGate identifies the critical turn in multi-turn dialogues where a response would complete hidden malicious intent, outperforming baselines on the new MTID dataset while keeping over-refusal low.
-
Evolve the Method, Not the Prompts: Evolutionary Synthesis of Jailbreak Attacks on LLMs
EvoSynth evolves code-based jailbreak algorithms via multi-agent self-correction, reaching 85.5% ASR on Claude-Sonnet-4.5 and 95.9% average across targets with greater diversity.
-
A Red Teaming Framework for Large Language Models: A Case Study on Faithfulness Evaluation
A multi-role red-teaming framework with attacker, target, and jury LLMs measures faithfulness in English and Arabic, finding false-premise prompts and length limits change unfaithfulness rates.
-
ADR: An Agentic Detection System for Enterprise Agentic AI Security
ADR is a three-component detection system for AI agents that combines telemetry sensors, red teaming, and two-tier detection, achieving 97.2% precision in a ten-month Uber deployment and outperforming baselines on the new ADR-Bench.