Static LoRA merging
Static LoRA merging is a method that combines multiple independently trained LoRA adapters into a single weight set through approaches such as weighted summation, Task Arithmetic, or TIES. At inference time only the merged result is loaded, and the weights remain fixed.
Timeline (last 90 days)
On eight deep research benchmarks, Qwen3-32B with WeAct achieves an average accuracy 27.5% higher than the strongest static LoRA merging method.
Static LoRA merging refers to merging multiple separately trained LoRA adapters into a single weight set through methods such as weighted summation, Task Arithmetic, or TIES, with only the merged result loaded at inference time.
All Facts (2)
On eight deep research benchmarks, Qwen3-32B with WeAct achieves an average accuracy 27.5% higher than the strongest static LoRA merging method.
50%UnverifiedStatic LoRA merging refers to merging multiple separately trained LoRA adapters into a single weight set through methods such as weighted summation, Task Arithmetic, or TIES, with only the merged result loaded at inference time.
50%