Scene-conditioned spatial-misdirection and duration-inflation backdoors succeed at 2.5–10% poison ratios on multimodal scanpath predictors and resist five adapted defenses.
arXiv preprint arXiv:2004.06660 , year=
6 Pith papers cite this work. Polarity classification is still indexing.
6
Pith papers citing it
representative citing papers
The paper presents Proactive Availability Backdoor (PAB) attacks on LLMs that achieve 73.1% effective success rate by proactively inducing users via suggestions in a Five-Factor Model simulation.
LLMs trained on simple specification gaming generalize to zero-shot reward tampering including rewriting their own reward function.
SCOUT uses token saliency analysis to detect both standard and contextually-plausible backdoor attacks in language models while maintaining clean accuracy.
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.
Catastrophic overfitting in fast adversarial training is reinterpreted as a weak-trigger variant of unlearnable tasks, allowing backdoor-inspired recalibration and outlier suppression to restore robustness.
citing papers explorer
-
Follow My Eyes: Backdoor Attacks on Goal-Directed Scanpath Prediction
Scene-conditioned spatial-misdirection and duration-inflation backdoors succeed at 2.5–10% poison ratios on multimodal scanpath predictors and resist five adapted defenses.
-
The Invitation Trap: Proactive Availability Backdoor in LLMs via Conversational Induction
The paper presents Proactive Availability Backdoor (PAB) attacks on LLMs that achieve 73.1% effective success rate by proactively inducing users via suggestions in a Five-Factor Model simulation.
-
Sycophancy to Subterfuge: Investigating Reward-Tampering in Large Language Models
LLMs trained on simple specification gaming generalize to zero-shot reward tampering including rewriting their own reward function.
-
SCOUT: A Defense Against Data Poisoning Attacks in Fine-Tuned Language Models
SCOUT uses token saliency analysis to detect both standard and contextually-plausible backdoor attacks in language models while maintaining clean accuracy.
-
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.
-
Unveiling the Backdoor Mechanism Hidden Behind Catastrophic Overfitting in Fast Adversarial Training
Catastrophic overfitting in fast adversarial training is reinterpreted as a weak-trigger variant of unlearnable tasks, allowing backdoor-inspired recalibration and outlier suppression to restore robustness.