{"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-2603-28092","title":"InkDrop: Invisible Backdoor Attacks Against Dataset Condensation","arxiv_id":"2603.28092","date":"2026-03-30","proceeding":null,"authors":["He Yang","Dongyi Lv","Song Ma","Wei Xi","Zhi Wang","Hanlin Gu","Yajie Wang"],"abstract":"Dataset Condensation (DC) is a data-efficient learning paradigm that synthesizes small yet informative datasets, enabling models to match the performance of full-data training. However, recent work exposes a critical vulnerability of DC to backdoor attacks, where malicious patterns (\\textit{e.g.}, triggers) are implanted into the condensation dataset, inducing targeted misclassification on specific inputs. Existing attacks always prioritize attack effectiveness and model utility, overlooking the crucial dimension of stealthiness. To bridge this gap, we propose InkDrop, which enhances the imperceptibility of malicious manipulation without degrading attack effectiveness and model utility. InkDrop leverages the inherent uncertainty near model decision boundaries, where minor input perturbations can induce semantic shifts, to construct a stealthy and effective backdoor attack. Specifically, InkDrop first selects candidate samples near the target decision boundary that exhibit latent semantic affinity to the target class. It then learns instance-dependent perturbations constrained by perceptual and spatial consistency, embedding targeted malicious behavior into the condensed dataset. Extensive experiments across diverse datasets validate the overall effectiveness of InkDrop, demonstrating its ability to integrate adversarial intent into condensed datasets while preserving model utility and minimizing detectability. Our code is available at https://github.com/lvdongyi/InkDrop.","url_abs":"https://arxiv.org/abs/2603.28092","url_pdf":"https://arxiv.org/pdf/2603.28092","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":{"atlas_url":"https://app.syntology.ai/?focus=2603.28092","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2603.28092"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+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/lvdongyi/InkDrop","reach":null}],"summary":{"unverified":7},"by_repo_kind":{"found_in_text":{"samples":7,"ran":0,"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":7,"samples":[{"code_sha256_prefix":"8dcf0de6d12c6bd3","entry":"com_regularizer","repo":"lvdongyi/InkDrop","repo_kind":"found_in_text","path":"backdoors/edge_case_backdoor.py","file_url":"https://github.com/lvdongyi/InkDrop/blob/HEAD/backdoors/edge_case_backdoor.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":"8dcf0de6d12c6bd3"}},{"code_sha256_prefix":"52d0879536c3edd2","entry":"freq_regularizer","repo":"lvdongyi/InkDrop","repo_kind":"found_in_text","path":"backdoors/edge_case_backdoor.py","file_url":"https://github.com/lvdongyi/InkDrop/blob/HEAD/backdoors/edge_case_backdoor.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":"52d0879536c3edd2"}},{"code_sha256_prefix":"5c9cef1327c93c26","entry":"get_activation","repo":"lvdongyi/InkDrop","repo_kind":"found_in_text","path":"backdoors/doorping_backdoor.py","file_url":"https://github.com/lvdongyi/InkDrop/blob/HEAD/backdoors/doorping_backdoor.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":"5c9cef1327c93c26"}},{"code_sha256_prefix":"a7302febefcac5d8","entry":"get_border_mask","repo":"lvdongyi/InkDrop","repo_kind":"found_in_text","path":"backdoors/casev2_backdoor.py","file_url":"https://github.com/lvdongyi/InkDrop/blob/HEAD/backdoors/casev2_backdoor.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":"a7302febefcac5d8"}},{"code_sha256_prefix":"9dcbf20acfe9626c","entry":"get_clip_image","repo":"lvdongyi/InkDrop","repo_kind":"found_in_text","path":"backdoors/case_backdoor.py","file_url":"https://github.com/lvdongyi/InkDrop/blob/HEAD/backdoors/case_backdoor.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":"9dcbf20acfe9626c"}},{"code_sha256_prefix":"6cc61f0163a597cd","entry":"get_middle_output","repo":"lvdongyi/InkDrop","repo_kind":"found_in_text","path":"backdoors/doorping_backdoor.py","file_url":"https://github.com/lvdongyi/InkDrop/blob/HEAD/backdoors/doorping_backdoor.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":"6cc61f0163a597cd"}},{"code_sha256_prefix":"19caf14e04f481ff","entry":"l2_loss","repo":"lvdongyi/InkDrop","repo_kind":"found_in_text","path":"backdoors/casev2_backdoor.py","file_url":"https://github.com/lvdongyi/InkDrop/blob/HEAD/backdoors/casev2_backdoor.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":"19caf14e04f481ff"}}]},"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}