LoGo is a training-free framework that dynamically selects and merges LoRA adapters at the instance level using signals from a single forward pass to handle diverse tasks.
In Thirty-fifth Conference on Neural Information Pro- cessing Systems Datasets and Benchmarks Track (Round 1)
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2025 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
LoRA on the Go: Instance-level Dynamic LoRA Selection and Merging
LoGo is a training-free framework that dynamically selects and merges LoRA adapters at the instance level using signals from a single forward pass to handle diverse tasks.