TeleLoRA: Teleporting Model-Specific Alignment Across LLMs
Mitigating Trojans in Large Language Models (LLMs) is one of many tasks where alignment data is LLM specific, as different LLMs have different Trojan triggers and trigger behaviors to be removed. In this paper, we introduce TeleLoRA (Teleporting Low-Rank Adaptation), a novel framework that synergizes model-specific alignment data across multiple LLMs to enable zero-shot Trojan mitigation on unseen LLMs without alignment data. TeleLoRA learns a unified generator of LoRA adapter weights by leveraging local activation information across multiple LLMs. This generator is designed to be permutation symmetric to generalize across models with different architectures and sizes. We optimize the model design for memory efficiency, making it feasible to learn with large-scale LLMs with minimal computational resources. Experiments on LLM Trojan mitigation benchmarks demonstrate that TeleLoRA effectively reduces attack success rates while preserving the benign performance of the models.
Code (0)
등록된 구현이 없습니다.
Tasks
modelMethods 이 논문이 사용한 방법론
Similar Papers 제목 키워드 기반
ABRA: Teleporting Fine-Tuned Knowledge Across Domains for Open-Vocabulary Object Detection
Although recent Open-Vocabulary Object Detection architectures, such as Grounding DINO, demonstrate strong zero-shot capabilities, their performance degrades significantly under domain shifts. Moreover, many domains of p…
Object DetectionNeural Fields as World Models
Humans rehearse possible futures offline, as in mental practice and perhaps dreaming, suggesting that world models may support task learning away from the environment. Standard machine learning world models compress visu…
AR-MAP: Are Autoregressive Large Language Models Implicit Teachers for Diffusion Large Language Models?
Diffusion Large Language Models (DLLMs) have emerged as a powerful alternative to autoregressive models, enabling parallel token generation across multiple positions. However, preference alignment of DLLMs remains challe…
Transfer LearningImproving the Distributional Alignment of LLMs using Supervision
The ability to accurately align LLMs with diverse population groups on subjective questions would have great value. In this work, we show that adding simple supervision can more consistently improve the alignment of LLM-…
Improving Adversarial Transferability in MLLMs via Dynamic Vision-Language Alignment Attack
Multimodal Large Language Models (MLLMs), built upon LLMs, have recently gained attention for their capabilities in image recognition and understanding. However, while MLLMs are vulnerable to adversarial attacks, the tra…