{"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/ab-using-images-and-sounds-for-indirect","title":"Abusing Images and Sounds for Indirect Instruction Injection in Multi-Modal LLMs","arxiv_id":"2307.10490","date":"2023-07-19","proceeding":null,"authors":["Eugene Bagdasaryan","Tsung-Yin Hsieh","Ben Nassi","Vitaly Shmatikov"],"abstract":"We demonstrate how images and sounds can be used for indirect prompt and instruction injection in multi-modal LLMs. An attacker generates an adversarial perturbation corresponding to the prompt and blends it into an image or audio recording. When the user asks the (unmodified, benign) model about the perturbed image or audio, the perturbation steers the model to output the attacker-chosen text and/or make the subsequent dialog follow the attacker's instruction. We illustrate this attack with several proof-of-concept examples targeting LLaVa and PandaGPT.","url_abs":"https://arxiv.org/abs/2307.10490v4","url_pdf":"https://arxiv.org/pdf/2307.10490v4.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"ab-using-images-and-sounds-for-indirect","repo_url":"https://github.com/ebagdasa/multimodal_injection","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2307.10490","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2307.10490"}},"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":[{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/ebagdasa/multimodal_injection","reach":{"status":"ok","spdx":"MIT"}}],"summary":{"ran_draft_wrong":2,"ran":1,"ran_honours":1},"by_repo_kind":{"official":{"samples":4,"ran":4,"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":"42a46570620cd9fa","entry":"get_chunk","repo":"ebagdasa/multimodal_injection","repo_kind":"official","path":"llava/eval/model_vqa.py","file_url":"https://github.com/ebagdasa/multimodal_injection/blob/HEAD/llava/eval/model_vqa.py","link_basis":"harvester_set","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"MISDECLARED","metamorphic_tier":"well_formed","behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"42a46570620cd9fa"}},{"code_sha256_prefix":"9d92b6748463e13b","entry":"inverse_normalize","repo":"ebagdasa/multimodal_injection","repo_kind":"official","path":"pandagpt/pandagpt_injection.py","file_url":"https://github.com/ebagdasa/multimodal_injection/blob/HEAD/pandagpt/pandagpt_injection.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"9d92b6748463e13b"}},{"code_sha256_prefix":"9b3c1cb391672ccb","entry":"load_image","repo":"ebagdasa/multimodal_injection","repo_kind":"official","path":"llava/llava_injection.py","file_url":"https://github.com/ebagdasa/multimodal_injection/blob/HEAD/llava/llava_injection.py","link_basis":"harvester_set","language":"python","status":"ran_honours","verification_level":1,"contract_check":"HONOURS","metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"9b3c1cb391672ccb"}},{"code_sha256_prefix":"076c252c52cbb161","entry":"split_list","repo":"ebagdasa/multimodal_injection","repo_kind":"official","path":"llava/eval/model_vqa.py","file_url":"https://github.com/ebagdasa/multimodal_injection/blob/HEAD/llava/eval/model_vqa.py","link_basis":"harvester_set","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"076c252c52cbb161"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}