MasqLoRA shows that an independent LoRA adapter can be trained on a few trigger-target pairs to backdoor diffusion models with 99.8% success rate while remaining stealthy when the trigger is absent.
LoRATK: LoRA Once, Backdoor Everywhere in the Share-and-Play Ecosystem
5 Pith papers cite this work. Polarity classification is still indexing.
5
Pith papers citing it
citation-role summary
background 1
citation-polarity summary
roles
background 1polarities
background 1representative citing papers
PoisonLoRA demonstrates ~100% attack success rates for stealthy LoRA poisoning via concept hijacking and task injection on real platforms, with robustness to base model transfer and multiple remixes.
Sparse autoencoders identify shared latent features across diverse backdoor attacks in LLMs that enable unified detection via classifiers, causal control via steering, and mitigation via ablation fine-tuning.
A Dirichlet-prior Bayesian estimator for model success probability replaces Pass@k, delivering faster-converging and more stable rankings with credible intervals on math benchmarks.
The paper introduces a new taxonomy for model merging methods and reviews their applications in LLMs, MLLMs, continual learning, multi-task learning, and other subfields while outlining open challenges.
citing papers explorer
-
When LoRA Betrays: Backdooring Text-to-Image Models by Masquerading as Benign Adapters
MasqLoRA shows that an independent LoRA adapter can be trained on a few trigger-target pairs to backdoor diffusion models with 99.8% success rate while remaining stealthy when the trigger is absent.
-
Customization under Fire: Plugin Poisoning in Text-to-Image Ecosystem
PoisonLoRA demonstrates ~100% attack success rates for stealthy LoRA poisoning via concept hijacking and task injection on real platforms, with robustness to base model transfer and multiple remixes.
-
Shared Latent Structures Enable Unified Backdoor Detection and Mitigation in LLMs
Sparse autoencoders identify shared latent features across diverse backdoor attacks in LLMs that enable unified detection via classifiers, causal control via steering, and mitigation via ablation fine-tuning.
-
Don't Pass@k: A Bayesian Framework for Large Language Model Evaluation
A Dirichlet-prior Bayesian estimator for model success probability replaces Pass@k, delivering faster-converging and more stable rankings with credible intervals on math benchmarks.
-
Model Merging in LLMs, MLLMs, and Beyond: Methods, Theories, Applications and Opportunities
The paper introduces a new taxonomy for model merging methods and reviews their applications in LLMs, MLLMs, continual learning, multi-task learning, and other subfields while outlining open challenges.