LoKI: Low-damage Knowledge Implanting of Large Language Models

Code

• github.com/Nexround/LoKI
  OfficialIn paperpytorch★ 18

Abstract

Fine-tuning adapts pretrained models for specific tasks but poses the risk of catastrophic forgetting (CF), where critical knowledge from pre-training is overwritten. Current Parameter-Efficient Fine-Tuning (PEFT) methods for Large Language Models (LLMs), while efficient, often sacrifice general capabilities. To address the issue of CF in a general-purpose PEFT framework, we propose Low-damage Knowledge Implanting (LoKI), a PEFT technique that is based on a mechanistic understanding of how knowledge is stored in transformer architectures. In two real-world scenarios, LoKI demonstrates task-specific performance that is comparable to or even surpasses that of full fine-tuning and LoRA-based methods across various model types, while significantly better preserving general capabilities. Our work connects mechanistic insights into LLM knowledge storage with practical fine-tuning objectives, achieving state-of-the-art trade-offs between task specialization and the preservation of general capabilities. Our implementation is publicly available as ready-to-use codehttps://github.com/Nexround/LoKI.

Tasks

Reproductions