REVIEW 14 cited by
Explore, Establish, Exploit: Red Teaming Language Models from Scratch
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
read the original abstract
Deploying large language models (LMs) can pose hazards from harmful outputs such as toxic or false text. Prior work has introduced automated tools that elicit harmful outputs to identify these risks. While this is a valuable step toward securing models, these approaches rely on a pre-existing way to efficiently classify undesirable outputs. Using a pre-existing classifier does not allow for red-teaming to be tailored to the target model. Furthermore, when failures can be easily classified in advance, red-teaming has limited marginal value because problems can be avoided by simply filtering training data and/or model outputs. Here, we consider red-teaming "from scratch," in which the adversary does not begin with a way to classify failures. Our framework consists of three steps: 1) Exploring the model's range of behaviors in the desired context; 2) Establishing a definition and measurement for undesired behavior (e.g., a classifier trained to reflect human evaluations); and 3) Exploiting the model's flaws using this measure to develop diverse adversarial prompts. We use this approach to red-team GPT-3 to discover classes of inputs that elicit false statements. In doing so, we construct the CommonClaim dataset of 20,000 statements labeled by humans as common-knowledge-true, common knowledge-false, or neither. We are making code and data available.
Forward citations
Cited by 14 Pith papers
-
Tailored untruths: How personalisation challenges LLM safeguards
A 1.6-million-text study of eight LLMs in four languages finds that adding demographic personae to disinformation prompts raises jailbreak rates from 78% to 82%.
-
Exposure is not manifestation: measurement target and output resolution jointly determine which behavioural-faithfulness evaluator wins
Small hyperbolic models (146M–3B) report 100% creative-seed preference, 90.7% compliance-gap detection, and a selective-gating skeleton–wallpaper memory pilot as a companion-AI stack.
-
A Systematic Investigation of RL-Jailbreaking in LLMs
Systematic investigation reveals that dense rewards and extended episode lengths primarily drive the success of RL jailbreaking in LLMs.
-
Just Keep Prompting: Evaluating Repetitive Socratic Prompting in VLMs
Repeated conversational pressure destabilizes VLM video answers — correct answers regress, wrong ones sometimes recover, and models differ sharply in how they respond.
-
FORTRESS: Frontier Risk Evaluation for National Security and Public Safety
A new benchmark with instance-specific rubrics measures frontier LLMs' willingness to assist with national security and public safety threats, alongside a paired over-refusal test.
-
MoGU V2: Toward a Higher Pareto Frontier Between Model Usability and Security
MoGUv2 embeds small routers in the deeper layers of LLMs to dynamically blend a helpful variant and a refusal variant, improving safety against jailbreak and fine-tuning attacks while preserving usability.
-
Anchoring Refusal Direction: Mitigating Safety Risks in Tuning via Projection Constraint
ProCon anchors each sample's hidden-state projection onto the LLM's initial refusal direction during instruction fine-tuning, reducing refusal-direction drift and safety risks with limited task-performance loss.
-
Quantized but Deceptive? A Multi-Dimensional Truthfulness Evaluation of Quantized LLMs
The study introduces TruthfulnessEval and reports that 4-bit quantization preserves simple true/false accuracy, but explicit 'lie' prompts make quantized and full-precision LLMs output falsehoods even when internal pr...
-
RedRFT: A Light-Weight Benchmark for Reinforcement Fine-Tuning-Based Red Teaming
RedRFT is a new open-source benchmark with a unified PPO backbone, five reimplemented red teaming baselines, a proposed diversity metric, and ablation insights.
-
SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models
Across 510 HarmBench behaviors and seven attack methods, GPT-4 models show more consistent jailbreak resilience than DeepSeek models, whose vulnerability grows with scale.
-
Adversarial Preference Learning for Robust LLM Alignment
APL iteratively trains an attacker to generate adversarial prompt rewrites and a defender to resist them, using the defender's own preference probabilities as the attack signal.
-
Guardians and Offenders: A Survey on Harmful Content Generation and Safety Mitigation of LLM
The submission's abstract promises an LLM safety survey, but the provided body is the opening page of an unrelated arithmetic-dynamics paper, so the artifact is internally inconsistent.
-
PRM-Free Security Alignment of Large Models via Red Teaming and Adversarial Training
A PRM-free alignment pipeline combining genetic algorithm red teaming and multi-objective adversarial training is claimed to beat PRM-based methods at 61% lower cost, but the experiments are unverifiable.
- Kaleidoscopic Teaming in Multi Agent Simulations
Discussion (0). Sign in to comment.