IDANI: Inference-time Domain Adaptation via Neuron-level Interventions
This addresses domain shift issues for users of pre-trained models, offering an efficient inference-time solution, though it is incremental as it builds on existing domain adaptation concepts.
The paper tackles the problem of domain adaptation for large pre-trained models by proposing a method that modifies test examples via neuron-level interventions to create counterfactual examples from the source domain, improving performance on unseen domains.
Large pre-trained models are usually fine-tuned on downstream task data, and tested on unseen data. When the train and test data come from different domains, the model is likely to struggle, as it is not adapted to the test domain. We propose a new approach for domain adaptation (DA), using neuron-level interventions: We modify the representation of each test example in specific neurons, resulting in a counterfactual example from the source domain, which the model is more familiar with. The modified example is then fed back into the model. While most other DA methods are applied during training time, ours is applied during inference only, making it more efficient and applicable. Our experiments show that our method improves performance on unseen domains.