{"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/2408-02544","title":"Caution for the Environment: Multimodal Agents are Susceptible to Environmental Distractions","arxiv_id":"2408.02544","date":"2024-08-05","proceeding":null,"authors":["Xinbei Ma","Yiting Wang","Yao Yao","Tongxin Yuan","Aston Zhang","Zhuosheng Zhang","Hai Zhao"],"abstract":"This paper investigates the faithfulness of multimodal large language model (MLLM) agents in the graphical user interface (GUI) environment, aiming to address the research question of whether multimodal GUI agents can be distracted by environmental context. A general setting is proposed where both the user and the agent are benign, and the environment, while not malicious, contains unrelated content. A wide range of MLLMs are evaluated as GUI agents using our simulated dataset, following three working patterns with different levels of perception. Experimental results reveal that even the most powerful models, whether generalist agents or specialist GUI agents, are susceptible to distractions. While recent studies predominantly focus on the helpfulness (i.e., action accuracy) of multimodal agents, our findings indicate that these agents are prone to environmental distractions, resulting in unfaithful behaviors. Furthermore, we switch to the adversarial perspective and implement environment injection, demonstrating that such unfaithfulness can be exploited, leading to unexpected risks.","url_abs":"https://arxiv.org/abs/2408.02544v1","url_pdf":"https://arxiv.org/pdf/2408.02544v1.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":"2408-02544","repo_url":"https://github.com/xbmxb/EnvDistraction","is_official":1,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok"}}],"tasks":[{"task_slug":"language-modeling","task_name":"Language Modeling"},{"task_slug":"language-modelling","task_name":"Language Modelling"},{"task_slug":"large-language-model","task_name":"Large Language Model"},{"task_slug":"multimodal-large-language-model","task_name":"Multimodal Large Language Model"}],"methods":[{"method_slug":"focus","method_name":"Focus"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2408.02544","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2408.02544"}},"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/xbmxb/EnvDistraction","reach":{"status":"ok"}}],"summary":{"ran":14,"unverified":2},"by_repo_kind":{"official":{"samples":16,"ran":14,"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":16,"samples":[{"code_sha256_prefix":"84f82a0ff9205c7b","entry":"add_commas","repo":"xbmxb/EnvDistraction","repo_kind":"official","path":"google_api.py","file_url":"https://github.com/xbmxb/EnvDistraction/blob/HEAD/google_api.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":"84f82a0ff9205c7b"}},{"code_sha256_prefix":"073de53b1531f4b6","entry":"check_answer","repo":"xbmxb/EnvDistraction","repo_kind":"official","path":"evaluation.py","file_url":"https://github.com/xbmxb/EnvDistraction/blob/HEAD/evaluation.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":"073de53b1531f4b6"}},{"code_sha256_prefix":"c44ec2a8c58bfa99","entry":"deduplicate_remove","repo":"xbmxb/EnvDistraction","repo_kind":"official","path":"amazon_products.py","file_url":"https://github.com/xbmxb/EnvDistraction/blob/HEAD/amazon_products.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":"c44ec2a8c58bfa99"}},{"code_sha256_prefix":"274a760ee5920652","entry":"deduplicate_string","repo":"xbmxb/EnvDistraction","repo_kind":"official","path":"amazon_products.py","file_url":"https://github.com/xbmxb/EnvDistraction/blob/HEAD/amazon_products.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":"274a760ee5920652"}},{"code_sha256_prefix":"f39135c9ca90db18","entry":"expand_exp","repo":"xbmxb/EnvDistraction","repo_kind":"official","path":"call_agents_0711cot.py","file_url":"https://github.com/xbmxb/EnvDistraction/blob/HEAD/call_agents_0711cot.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":"f39135c9ca90db18"}},{"code_sha256_prefix":"2e87e58f96ce5d53","entry":"format_tokens_llama","repo":"xbmxb/EnvDistraction","repo_kind":"official","path":"format_tokens.py","file_url":"https://github.com/xbmxb/EnvDistraction/blob/HEAD/format_tokens.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":"2e87e58f96ce5d53"}},{"code_sha256_prefix":"0c293a4793725f7c","entry":"format_tokens_mistral","repo":"xbmxb/EnvDistraction","repo_kind":"official","path":"format_tokens.py","file_url":"https://github.com/xbmxb/EnvDistraction/blob/HEAD/format_tokens.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":"0c293a4793725f7c"}},{"code_sha256_prefix":"3bd73a3bbe003219","entry":"format_tokens_yi","repo":"xbmxb/EnvDistraction","repo_kind":"official","path":"format_tokens.py","file_url":"https://github.com/xbmxb/EnvDistraction/blob/HEAD/format_tokens.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":"3bd73a3bbe003219"}},{"code_sha256_prefix":"2075373a61e75fd7","entry":"has_answer","repo":"xbmxb/EnvDistraction","repo_kind":"official","path":"evaluation.py","file_url":"https://github.com/xbmxb/EnvDistraction/blob/HEAD/evaluation.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":"2075373a61e75fd7"}},{"code_sha256_prefix":"0f449824dedf80a7","entry":"hits","repo":"xbmxb/EnvDistraction","repo_kind":"official","path":"evaluation.py","file_url":"https://github.com/xbmxb/EnvDistraction/blob/HEAD/evaluation.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":"0f449824dedf80a7"}},{"code_sha256_prefix":"2e99ae3171ad63cd","entry":"remove_leading_zeros","repo":"xbmxb/EnvDistraction","repo_kind":"official","path":"call_agents_0711cot.py","file_url":"https://github.com/xbmxb/EnvDistraction/blob/HEAD/call_agents_0711cot.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":"2e99ae3171ad63cd"}},{"code_sha256_prefix":"11bdac3ade81f187","entry":"resize_encode_image","repo":"xbmxb/EnvDistraction","repo_kind":"official","path":"annotation.py","file_url":"https://github.com/xbmxb/EnvDistraction/blob/HEAD/annotation.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":"11bdac3ade81f187"}},{"code_sha256_prefix":"46b4a889473dc3da","entry":"sample_counting","repo":"xbmxb/EnvDistraction","repo_kind":"official","path":"annotation.py","file_url":"https://github.com/xbmxb/EnvDistraction/blob/HEAD/annotation.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":"46b4a889473dc3da"}},{"code_sha256_prefix":"c3586cb845bba668","entry":"to_html","repo":"xbmxb/EnvDistraction","repo_kind":"official","path":"google_api.py","file_url":"https://github.com/xbmxb/EnvDistraction/blob/HEAD/google_api.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":"c3586cb845bba668"}},{"code_sha256_prefix":"67eccaf3e4aa4356","entry":"auto_utter","repo":"xbmxb/EnvDistraction","repo_kind":"official","path":"autorepyly.py","file_url":"https://github.com/xbmxb/EnvDistraction/blob/HEAD/autorepyly.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":"67eccaf3e4aa4356"}},{"code_sha256_prefix":"dc63deb0fb0338e8","entry":"resize_encode_image","repo":"xbmxb/EnvDistraction","repo_kind":"official","path":"call_agents_0711cot.py","file_url":"https://github.com/xbmxb/EnvDistraction/blob/HEAD/call_agents_0711cot.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":"dc63deb0fb0338e8"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}