{"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-2603-29410","title":"AGFT: Alignment-Guided Fine-Tuning for Zero-Shot Adversarial Robustness of Vision-Language Models","arxiv_id":"2603.29410","date":"2026-03-31","proceeding":null,"authors":["Yubo Cui","Xianchao Guan","Zijun Xiong","Zheng Zhang"],"abstract":"Pre-trained vision-language models (VLMs) exhibit strong zero-shot generalization but remain vulnerable to adversarial perturbations. Existing classification-guided adversarial fine-tuning methods often disrupt pre-trained cross-modal alignment, weakening visual-textual correspondence and degrading zero-shot performance. In this paper, we propose an Alignment-Guided Fine-Tuning (AGFT) framework that enhances zero-shot adversarial robustness while preserving the cross-modal semantic structure. Unlike label-based methods that rely on hard labels and fail to maintain the relative relationships between image and text, AGFT leverages the probabilistic predictions of the original model for text-guided adversarial training, which aligns adversarial visual features with textual embeddings via soft alignment distributions, improving zero-shot adversarial robustness. To address structural discrepancies introduced by fine-tuning, we introduce a distribution consistency calibration mechanism that adjusts the robust model output to match a temperature-scaled version of the pre-trained model predictions. Extensive experiments across multiple zero-shot benchmarks demonstrate that AGFT outperforms state-of-the-art methods while significantly improving zero-shot adversarial robustness.","url_abs":"https://arxiv.org/abs/2603.29410","url_pdf":"https://arxiv.org/pdf/2603.29410","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=2603.29410","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2603.29410"}},"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/YuboCui/AGFT","reach":null}],"summary":{"unverified":6},"by_repo_kind":{"found_in_text":{"samples":6,"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":6,"samples":[{"code_sha256_prefix":"cd82db2984e88662","entry":"get_train_dataset","repo":"YuboCui/AGFT","repo_kind":"found_in_text","path":"utils.py","file_url":"https://github.com/YuboCui/AGFT/blob/HEAD/utils.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":"cd82db2984e88662"}},{"code_sha256_prefix":"3a72468987fa3206","entry":"get_val_attack","repo":"YuboCui/AGFT","repo_kind":"found_in_text","path":"utils.py","file_url":"https://github.com/YuboCui/AGFT/blob/HEAD/utils.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":"3a72468987fa3206"}},{"code_sha256_prefix":"5ed58d803dcaec0f","entry":"get_val_dataset","repo":"YuboCui/AGFT","repo_kind":"found_in_text","path":"utils.py","file_url":"https://github.com/YuboCui/AGFT/blob/HEAD/utils.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":"5ed58d803dcaec0f"}},{"code_sha256_prefix":"523d7c2de2a69ace","entry":"load_folder2name","repo":"YuboCui/AGFT","repo_kind":"found_in_text","path":"clip_dataset.py","file_url":"https://github.com/YuboCui/AGFT/blob/HEAD/clip_dataset.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":"523d7c2de2a69ace"}},{"code_sha256_prefix":"801704a2c8af559e","entry":"refine_classname","repo":"YuboCui/AGFT","repo_kind":"found_in_text","path":"clip_dataset.py","file_url":"https://github.com/YuboCui/AGFT/blob/HEAD/clip_dataset.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":"801704a2c8af559e"}},{"code_sha256_prefix":"e34f1b1db45c292d","entry":"validate_one","repo":"YuboCui/AGFT","repo_kind":"found_in_text","path":"validate.py","file_url":"https://github.com/YuboCui/AGFT/blob/HEAD/validate.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":"e34f1b1db45c292d"}}]},"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":{"kind":"leaderboard_placements","source":"Syntology's leaderboard-shaped extractor over the paper's own arXiv-HTML tables: a model pointed at a cell, the number was read from that cell and checked against the board's metric, dataset, split and scale, and an independent check accepted the entry; not reviewed by the paper's authors or the archive's editors","extractor_model":"global.anthropic.claude-sonnet-4-5-20250929-v1:0","verifier_model":null,"prompt_sha":"fa63d4bb9d755694","coverage":{"sentence":"Syntology has checked 6,264 of the 9,581 papers on this site that are newer than the archive; results from the others appear after they are checked.","papers_newer_than_archive":9581,"papers_checked":6264},"entries":[],"not_placed":{"boards":0,"rejected_by_independent_check":0,"refused_by_a_rule":0,"check_did_not_answer":0,"proposed_without_a_cell":0,"declined_by_site":0}}}