Activated LoRA: Fine-tuned LLMs for Intrinsics
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866909822124818432 |
|---|---|
| author | Greenewald, Kristjan Lastras, Luis Parnell, Thomas Shah, Vraj Popa, Lucian Zizzo, Giulio Gunasekara, Chulaka Rawat, Ambrish Cox, David |
| author_facet | Greenewald, Kristjan Lastras, Luis Parnell, Thomas Shah, Vraj Popa, Lucian Zizzo, Giulio Gunasekara, Chulaka Rawat, Ambrish Cox, David |
| contents | Low-Rank Adaptation (LoRA) has emerged as a highly efficient framework for finetuning the weights of large foundation models, and has become the go-to method for data-driven customization of LLMs. Despite the promise of highly customized behaviors and capabilities, switching between relevant LoRAs in a multiturn setting is inefficient, as the key-value (KV) cache of the entire turn history must be recomputed with the LoRA weights before generation can begin. To address this problem, we propose Activated LoRA (aLoRA), an adapter architecture which modifies the LoRA framework to only adapt weights for the tokens in the sequence after the aLoRA is invoked. This change crucially allows aLoRA to accept the base model's KV cache of the input string, meaning that aLoRA can be instantly activated whenever needed in a chain without recomputing the prior keys and values. This enables building what we call intrinsics, i.e. specialized models invoked to perform well-defined operations on portions of an input chain or conversation that otherwise uses the base model by default. We train a set of aLoRA-based intrinsics models, demonstrating competitive accuracy with standard LoRA while significantly improving inference efficiency. We contributed our Activated LoRA implementation to the Huggingface PEFT library https://github.com/huggingface/peft. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2504_12397 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Activated LoRA: Fine-tuned LLMs for Intrinsics Greenewald, Kristjan Lastras, Luis Parnell, Thomas Shah, Vraj Popa, Lucian Zizzo, Giulio Gunasekara, Chulaka Rawat, Ambrish Cox, David Machine Learning Artificial Intelligence Low-Rank Adaptation (LoRA) has emerged as a highly efficient framework for finetuning the weights of large foundation models, and has become the go-to method for data-driven customization of LLMs. Despite the promise of highly customized behaviors and capabilities, switching between relevant LoRAs in a multiturn setting is inefficient, as the key-value (KV) cache of the entire turn history must be recomputed with the LoRA weights before generation can begin. To address this problem, we propose Activated LoRA (aLoRA), an adapter architecture which modifies the LoRA framework to only adapt weights for the tokens in the sequence after the aLoRA is invoked. This change crucially allows aLoRA to accept the base model's KV cache of the input string, meaning that aLoRA can be instantly activated whenever needed in a chain without recomputing the prior keys and values. This enables building what we call intrinsics, i.e. specialized models invoked to perform well-defined operations on portions of an input chain or conversation that otherwise uses the base model by default. We train a set of aLoRA-based intrinsics models, demonstrating competitive accuracy with standard LoRA while significantly improving inference efficiency. We contributed our Activated LoRA implementation to the Huggingface PEFT library https://github.com/huggingface/peft. |
| title | Activated LoRA: Fine-tuned LLMs for Intrinsics |
| topic | Machine Learning Artificial Intelligence |
| url | https://arxiv.org/abs/2504.12397 |