By sharing the B matrix across adapters instead of the A matrix, ALoRA and Fed-ALoRA deliver more balanced performance in multi-task and federated LLM fine-tuning.
10.0% CloQA 40.1% GenQA50.0% IE 24.8% OpnQA 25.3% Figure 7: Data distribution of clients
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Rethinking Parameter Sharing for LLM Fine-Tuning with Multiple LoRAs
By sharing the B matrix across adapters instead of the A matrix, ALoRA and Fed-ALoRA deliver more balanced performance in multi-task and federated LLM fine-tuning.