AI Paper Atlas

Research map / Label-efficient and robust learning / Domain adaptation and generalization

Holistic Transfer: Towards Non-Disruptive Fine-Tuning with Partial Target Data

Cheng-Hao Tu, Hong-You Chen, Zheda Mai, Jike Zhong, Vardaan Pahuja, Tanya Berger-Wolf, Song Gao, Charles Stewart, Yu Su, Wei-Lun Chao

NeurIPS 2023 · Label-efficient and robust learning › Domain adaptation and generalization · approach: Statistical transfer learning · 1 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
freeze the linear classifier during fine-tuning (ablation)Naive Target / SGD · Office-Home Avgunseen accuracy (%): 22.36 -> 43.44Improved
add selective distillation loss on unseen logits (ablation)SGD (frozen) · Office-Home Avgunseen accuracy (%): 43.44 -> 57.41Improved
add feature rank regularization (ablation)SGD (frozen) · Office-Home Avgunseen accuracy (%): 43.44 -> 55.64Improved
replace SGD with leave-out local SGD (LOLSGD) (ablation)SGD (frozen) · Office-Home Avgunseen accuracy (%): 43.44 -> 53.98Improved
replace SWAD with LOLSGD (ablation)SWAD (frozen) · Office-Home Avgunseen accuracy (%): 45.30 -> 53.98Improved
add distillation and rank regularizers to LOLSGD (ablation)LOLSGD (frozen) · Office-Home Avgunseen accuracy (%): 53.98 -> 66.54Improved
update only batchnorm statistics instead of naive fine-tuningNaive Target · Office-Home Avgunseen accuracy (%): 22.36 -> 67.51Improved
apply LP-FT instead of naive fine-tuningNaive Target · Office-Home Avgunseen accuracy (%): 22.36 -> 15.60Hurt
ensemble predictions with source model (SE) (ablation)Naive Target · Office-Home Avgoverall accuracy (%): 53.50 -> 67.46Improved
add source ensemble to LOLSGD + distill (ablation)LOLSGD+distill · FEMNISToverall accuracy (%): 85.76 -> 85.10Hurt
replace SGD with LOLSGD (ablation)SGD (frozen) · FEMNISTunseen accuracy (%): 64.00 -> 66.27Mixed
replace SGD with LOLSGD (ablation)SGD (frozen) · iWildCamunseen accuracy (%): 7.91 -> 13.30Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy