{"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-2605-24816","title":"AOEPT: Breaking the Implicit Modality-Reduction Bottleneck in Modality-Missing Prompt Tuning","arxiv_id":"2605.24816","date":"2026-05-24","proceeding":"ICML","authors":["Jian Lang","Rongpei Hong","Ting Zhong","Fan Zhou"],"abstract":"Deploying multimodal systems in real-world environments often entails handling modality-missing scenarios, where one or more modalities are unavailable. While recent studies address this challenge for the general Multimodal Transformer (MT) architecture via prompt tuning, we identify a fundamental limitation in these methods: the Implicit Modality-Reduction bottleneck. By conditioning prompts solely on the observed modalities, they inadvertently restrict the reasoning scope of MTs to the modality-reduced subspace, cutting off access to the latent information sources of the missing modalities. To overcome this limitation, we propose AOEPT, which pioneers a novel modal-contextualized prompting fashion. Specifically, we introduce lightweight Modal-Contextualized Prompts (MCPs) that distill global modality-wise priors from training data, serving as latent repositories of the information sources for missing modalities. Conditioned on the remaining modalities, these MCPs are instantiated into instance-aware prompts that selectively augment missing-modality information for each sample, thereby restoring the reasoning scope of MTs beyond the observed-modality-only subspace. Experiments across various multimodal benchmarks and backbones confirm the strong performance of AOEPT, with minimal computational overhead.","url_abs":"https://arxiv.org/abs/2605.24816","url_pdf":"https://arxiv.org/pdf/2605.24816","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=2605.24816","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2605.24816"}},"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/Jian-Lang/AOEPT","reach":null}],"summary":{"ran":1,"unverified":2},"by_repo_kind":{"found_in_text":{"samples":3,"ran":1,"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":0,"samples":[{"code_sha256_prefix":"84e2db3e4965bd1c","entry":"Base","repo":"Jian-Lang/AOEPT","repo_kind":"found_in_text","path":"core/model/AOEPT/AOEPT_CLIP_model.py","file_url":"https://github.com/Jian-Lang/AOEPT/blob/HEAD/core/model/AOEPT/AOEPT_CLIP_model.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"84e2db3e4965bd1c"}},{"code_sha256_prefix":"a2a93a7443ff08f3","entry":"AOEPT","repo":"Jian-Lang/AOEPT","repo_kind":"found_in_text","path":"core/model/AOEPT/AOEPT_CLIP_model.py","file_url":"https://github.com/Jian-Lang/AOEPT/blob/HEAD/core/model/AOEPT/AOEPT_CLIP_model.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"a2a93a7443ff08f3"}},{"code_sha256_prefix":"538390567e78fc24","entry":"init_weights","repo":"Jian-Lang/AOEPT","repo_kind":"found_in_text","path":"core/model/AOEPT/AOEPT_CLIP_model.py","file_url":"https://github.com/Jian-Lang/AOEPT/blob/HEAD/core/model/AOEPT/AOEPT_CLIP_model.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"538390567e78fc24"}}]},"arxiv_metadata":{"licence":"arXiv metadata, CC0 1.0 (https://info.arxiv.org/help/license)","fields":["title","abstract","authors","date"],"primary_category":"cs.CV","source":"arxiv_2026.jsonl"},"syntology_extracted_results":null}