{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/arxiv-2603-21928","title":"The Golden Subspace: Where Efficiency Meets Generalization in Continual Test-Time Adaptation","arxiv_id":"2603.21928","date":"2026-03-23","proceeding":null,"authors":["Guannan Lai","Da-Wei Zhou","Zhenguo Li","Han-Jia Ye"],"abstract":"Continual Test-Time Adaptation (CTTA) aims to enable models to adapt online to unlabeled data streams under distribution shift without accessing source data. Existing CTTA methods face an efficiency-generalization trade-off: updating more parameters improves adaptation but severely reduces online inference efficiency. An ideal solution is to achieve comparable adaptation with minimal feature updates; we call this minimal subspace the golden subspace. We prove its existence in a single-step adaptation setting and show that it coincides with the row space of the pretrained classifier. To enable online maintenance of this subspace, we introduce the sample-wise Average Gradient Outer Product (AGOP) as an efficient proxy for estimating the classifier weights without retraining. Building on these insights, we propose Guided Online Low-rank Directional adaptation (GOLD), which uses a lightweight adapter to project features onto the golden subspace and learns a compact scaling vector while the subspace is dynamically updated via AGOP. Extensive experiments on classification and segmentation benchmarks, including autonomous-driving scenarios, demonstrate that GOLD attains superior efficiency, stability, and overall performance. Our code is available at https://github.com/AIGNLAI/GOLD.","url_abs":"https://arxiv.org/abs/2603.21928","url_pdf":"https://arxiv.org/pdf/2603.21928","source":{"archive":null,"snapshot":"2025-07-28","note":"not in the Papers with Code archive (frozen at the snapshot)","row_kind":"graph","title_abstract_authors_date":"arXiv metadata, CC0 1.0 (https://info.arxiv.org/help/license)"},"code_links":[],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/2603.21928","atlas_url":"https://app.syntology.ai/?focus=2603.21928","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2603.21928"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-25T09:33:49+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"mentioned_in_github":null,"is_official":null,"provenance":"deterministic:regex_extraction","mentioned_in_paper":null,"url":"https://github.com/AIGNLAI/GOLD","reach":null}],"summary":{"ran":12,"unverified":1},"by_repo_kind":{"found_in_text":{"samples":13,"ran":12,"repositories":1}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":13,"samples":[{"code_sha256_prefix":"53b1e10b7cc3d6de","entry":"complete_data_dir_path","repo":"AIGNLAI/GOLD","repo_kind":"found_in_text","path":"classification/conf.py","file_url":"https://github.com/AIGNLAI/GOLD/blob/HEAD/classification/conf.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"53b1e10b7cc3d6de"}},{"code_sha256_prefix":"0c4495cb2187be84","entry":"create_file_list","repo":"AIGNLAI/GOLD","repo_kind":"found_in_text","path":"classification/datasets/imagenet_subsets.py","file_url":"https://github.com/AIGNLAI/GOLD/blob/HEAD/classification/datasets/imagenet_subsets.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"0c4495cb2187be84"}},{"code_sha256_prefix":"22a042db3ce6f8bb","entry":"ds_name2pytorch_ds_name","repo":"AIGNLAI/GOLD","repo_kind":"found_in_text","path":"classification/conf.py","file_url":"https://github.com/AIGNLAI/GOLD/blob/HEAD/classification/conf.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"22a042db3ce6f8bb"}},{"code_sha256_prefix":"ef670fd2920b7ce2","entry":"forward_decorator","repo":"AIGNLAI/GOLD","repo_kind":"found_in_text","path":"classification/methods/base.py","file_url":"https://github.com/AIGNLAI/GOLD/blob/HEAD/classification/methods/base.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"ef670fd2920b7ce2"}},{"code_sha256_prefix":"4d83f34a9766b4a9","entry":"get_num_classes","repo":"AIGNLAI/GOLD","repo_kind":"found_in_text","path":"classification/conf.py","file_url":"https://github.com/AIGNLAI/GOLD/blob/HEAD/classification/conf.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"4d83f34a9766b4a9"}},{"code_sha256_prefix":"b086cfc2dfe368f4","entry":"kernel","repo":"AIGNLAI/GOLD","repo_kind":"found_in_text","path":"classification/methods/cmf.py","file_url":"https://github.com/AIGNLAI/GOLD/blob/HEAD/classification/methods/cmf.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"b086cfc2dfe368f4"}},{"code_sha256_prefix":"6c5a8d4a69a975c2","entry":"refine_predictions","repo":"AIGNLAI/GOLD","repo_kind":"found_in_text","path":"classification/methods/adacontrast.py","file_url":"https://github.com/AIGNLAI/GOLD/blob/HEAD/classification/methods/adacontrast.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"6c5a8d4a69a975c2"}},{"code_sha256_prefix":"ba822e549de6a087","entry":"setup_adacontrast_optimizer","repo":"AIGNLAI/GOLD","repo_kind":"found_in_text","path":"classification/methods/adacontrast.py","file_url":"https://github.com/AIGNLAI/GOLD/blob/HEAD/classification/methods/adacontrast.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"ba822e549de6a087"}},{"code_sha256_prefix":"110b394c4d611e0b","entry":"soft_k_nearest_neighbors","repo":"AIGNLAI/GOLD","repo_kind":"found_in_text","path":"classification/methods/adacontrast.py","file_url":"https://github.com/AIGNLAI/GOLD/blob/HEAD/classification/methods/adacontrast.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"110b394c4d611e0b"}},{"code_sha256_prefix":"991d0f3401682804","entry":"softmax_entropy_cifar","repo":"AIGNLAI/GOLD","repo_kind":"found_in_text","path":"classification/methods/cotta.py","file_url":"https://github.com/AIGNLAI/GOLD/blob/HEAD/classification/methods/cotta.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"991d0f3401682804"}},{"code_sha256_prefix":"6bdfed740b3da7ce","entry":"softmax_entropy_imagenet","repo":"AIGNLAI/GOLD","repo_kind":"found_in_text","path":"classification/methods/cotta.py","file_url":"https://github.com/AIGNLAI/GOLD/blob/HEAD/classification/methods/cotta.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"6bdfed740b3da7ce"}},{"code_sha256_prefix":"073ab5af80ddbe36","entry":"update_model_probs","repo":"AIGNLAI/GOLD","repo_kind":"found_in_text","path":"classification/methods/cmf.py","file_url":"https://github.com/AIGNLAI/GOLD/blob/HEAD/classification/methods/cmf.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"073ab5af80ddbe36"}},{"code_sha256_prefix":"b723ea0fcdc2c948","entry":"create_imagenet_subset","repo":"AIGNLAI/GOLD","repo_kind":"found_in_text","path":"classification/datasets/imagenet_subsets.py","file_url":"https://github.com/AIGNLAI/GOLD/blob/HEAD/classification/datasets/imagenet_subsets.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"b723ea0fcdc2c948"}}]},"arxiv_metadata":{"licence":"arXiv metadata, CC0 1.0 (https://info.arxiv.org/help/license)","fields":["title","abstract","authors","date"],"primary_category":"cs.LG","source":"arxiv_2026.jsonl"},"syntology_extracted_results":null}